Search results for: prediction interval

Authors: Natalia Kartushina

Abstract:

Foreign (second) language (L2) learning is highly promoted in modern society. Students are encouraged to study abroad (SA) to achieve the most effective learning outcomes. However, L2 learning has side effects for native language (L1) production, as L1 sounds might show a drift from the L1 norms towards those of the L2, and this, even after a short period of L2 learning. L1 assimilatory drift has been attributed to a strong perceptual association between similar L1 and L2 sounds in the mind of L2 leaners; thus, a change in the production of an L2 target leads to the change in the production of the related L1 sound. However, nowadays, it is quite common that speakers acquire two languages from birth, as, for example, it is the case for many bilingual communities (e.g., Basque and Spanish in the Basque Country). Yet, it remains to be established how FL learning affects native production in individuals who have two native languages, i.e., in simultaneous or very early bilinguals. Does FL learning (here a third language, L3) affect bilinguals’ both languages or only one? What factors determine which of the bilinguals’ languages is more susceptible to change? The current study examines the effects of L3 (English) learning on the production of vowels in the two native languages of simultaneous Spanish-Basque bilingual adolescents enrolled into the Erasmus SA English program. Ten bilingual speakers read five Spanish and Basque consonant-vowel-consonant-vowel words two months before their SA and the next day after their arrival back to Spain. Each word contained the target vowel in the stressed syllable and was repeated five times. Acoustic analyses measuring vowel openness (F1) and backness (F2) were performed. Two possible outcomes were considered. First, we predicted that L3 learning would affect the production of only one language and this would be the language that would be used the most in contact with English during the SA period. This prediction stems from the results of recent studies showing that early bilinguals have separate phonological systems for each of their languages; and that late FL learner (as it is the case of our participants), who tend to use their L1 in language-mixing contexts, have more L2-accented L1 speech. The second possibility stated that L3 learning would affect both of the bilinguals’ languages in line with the studies showing that bilinguals’ L1 and L2 phonologies interact and constantly co-influence each other. The results revealed that speakers who used both languages equally often (balanced users) showed an F1 drift in both languages toward the F1 of the English vowel space. Unbalanced speakers, however, showed a drift only in the less used language. The results are discussed in light of recent studies suggesting that the amount of language use is a strong predictor of the authenticity in speech production with less language use leading to more foreign-accented speech and, eventually, to language attrition.

Keywords: language-contact, multilingualism, phonetic drift, bilinguals' production

Procedia PDF Downloads 110

Authors: Muhammad Abu-Khaizaran, Ahmad Faza’, Tariq Othman, Yahia Yousef

Abstract:

This paper presents a study to provide sufficient and reliable information about constructing a Photovoltaic energy profile of the Birzeit University campus (BZU) based on the weather forecast. The developed Photovoltaic energy profile helps to predict the energy yield of the Photovoltaic systems based on the weather forecast and hence helps planning energy production and consumption. Two models will be developed in this paper; a Clear Sky Irradiance model and a Cloud-Cover Radiation model to predict the irradiance for a clear sky day and a cloudy day, respectively. The adopted procedure for developing such models takes into consideration two levels of abstraction. First, irradiance and weather data were acquired by a sensory (measurement) system installed on the rooftop of the Information Technology College building at Birzeit University campus. Second, power readings of a fully operational 51kW commercial Photovoltaic system installed in the University at the rooftop of the adjacent College of Pharmacy-Nursing and Health Professions building are used to validate the output of a simulation model and to help refine its structure. Based on a comparison between a mathematical model, which calculates Clear Sky Irradiance for the University location and two sets of accumulated measured data, it is found that the simulation system offers an accurate resemblance to the installed PV power station on clear sky days. However, these comparisons show a divergence between the expected energy yield and actual energy yield in extreme weather conditions, including clouding and soiling effects. Therefore, a more accurate prediction model for irradiance that takes into consideration weather factors, such as relative humidity and cloudiness, which affect irradiance, was developed; Cloud-Cover Radiation Model (CRM). The equivalent mathematical formulas implement corrections to provide more accurate inputs to the simulation system. The results of the CRM show a very good match with the actual measured irradiance during a cloudy day. The developed Photovoltaic profile helps in predicting the output energy yield of the Photovoltaic system installed at the University campus based on the predicted weather conditions. The simulation and practical results for both models are in a very good match.

Keywords: clear-sky irradiance model, cloud-cover radiation model, photovoltaic, weather forecast

Procedia PDF Downloads 133

Authors: Gautier Viaud, Paul-Henry Cournède

Abstract:

Plant growth models have been used extensively for the prediction of the phenotypic performance of plants. However, they remain most often calibrated for a given genotype and therefore do not take into account genotype by environment interactions. One way of achieving such an objective is to consider Bayesian hierarchical models. Three levels can be identified in such models: The first level describes how a given growth model describes the phenotype of the plant as a function of individual parameters, the second level describes how these individual parameters are distributed within a plant population, the third level corresponds to the attribution of priors on population parameters. Thanks to the Bayesian framework, choosing appropriate priors for the population parameters permits to derive analytical expressions for the full conditional distributions of these population parameters. As plant growth models are of a nonlinear nature, individual parameters cannot be sampled explicitly, and a Metropolis step must be performed. This allows for the use of a hybrid Gibbs--Metropolis sampler. A generic approach was devised for the implementation of both general state space models and estimation algorithms within a programming platform. It was designed using the Julia language, which combines an elegant syntax, metaprogramming capabilities and exhibits high efficiency. Results were obtained for Arabidopsis thaliana on both simulated and real data. An organ-scale Greenlab model for the latter is thus presented, where the surface areas of each individual leaf can be simulated. It is assumed that the error made on the measurement of leaf areas is proportional to the leaf area itself; multiplicative normal noises for the observations are therefore used. Real data were obtained via image analysis of zenithal images of Arabidopsis thaliana over a period of 21 days using a two-step segmentation and tracking algorithm which notably takes advantage of the Arabidopsis thaliana phyllotaxy. Since the model formulation is rather flexible, there is no need that the data for a single individual be available at all times, nor that the times at which data is available be the same for all the different individuals. This allows to discard data from image analysis when it is not considered reliable enough, thereby providing low-biased data in large quantity for leaf areas. The proposed model precisely reproduces the dynamics of Arabidopsis thaliana’s growth while accounting for the variability between genotypes. In addition to the estimation of the population parameters, the level of variability is an interesting indicator of the genotypic stability of model parameters. A promising perspective is to test whether some of the latter should be considered as fixed effects.

Keywords: bayesian, genotypic differentiation, hierarchical models, plant growth models

Procedia PDF Downloads 304

Authors: Oluwatosin M. A. Jesuyon

Abstract:

In order to propose a simpler tool that eliminates the age-long problems associated with the traditional index method for selection of multiple traits in broilers, the Barttlet factor regression equation is being proposed as an alternative selection tool. 100 day-old chicks each of Arbor Acres (AA) and Annak (AN) broiler strains were obtained from two rival hatcheries in Ibadan Nigeria. These were raised in deep litter system in a 56-day feeding trial at the University of Ibadan Teaching and Research Farm, located in South-west Tropical Nigeria. The body weight and body dimensions were measured and recorded during the trial period. Eight (8) zoometric measurements namely live weight (g), abdominal circumference, abdominal length, breast width, leg length, height, wing length and thigh circumference (all in cm) were recorded randomly from 20 birds within strain, at a fixed time on the first day of the new week respectively with a 5-kg capacity Camry scale. These records were analyzed and compared using completely randomized design (CRD) of SPSS analytical software, with the means procedure, Factor Scores (FS) in stepwise Multiple Linear Regression (MLR) procedure for initial live weight equations. Bartlett Factor Score (BFS) analysis extracted 2 factors for each strain, termed Body-length and Thigh-meatiness Factors for AA, and; Breast Size and Height Factors for AN. These derived orthogonal factors assisted in deducing and comparing trait-combinations that best describe body conformation and Meatiness in experimental broilers. BFS procedure yielded different body conformational traits for the two strains, thus indicating the different economic traits and advantages of strains. These factors could be useful as selection criteria for improving desired economic traits. The final Bartlett Factor Regression equations for prediction of body weight were highly significant with P < 0.0001, R2 of 0.92 and above, VIF of 1.00, and DW of 1.90 and 1.47 for Arbor Acres and Annak respectively. These FSR equations could be used as a simple and potent tool for selection during poultry flock improvement, it could also be used to estimate selection index of flocks in order to discriminate between strains, and evaluate consumer preference traits in broilers.

Keywords: alternative selection tool, Bartlet factor regression model, consumer preference trait, linear and body measurements, live body weight

Procedia PDF Downloads 203

Authors: Sulemana Ibrahim

Abstract:

Food security remains a paramount global challenge, with vulnerable regions grappling with issues of hunger and malnutrition. This study embarks on a comprehensive exploration of data-driven strategies aimed at ameliorating food security in such regions. Our research employs a multifaceted approach, integrating data analytics to predict crop yields, optimizing supply chains, and enhancing food distribution networks. The study unfolds as a multi-dimensional analysis, commencing with the development of robust machine learning models harnessing remote sensing data, historical crop yield records, and meteorological data to foresee crop yields. These predictive models, underpinned by convolutional and recurrent neural networks, furnish critical insights into anticipated harvests, empowering proactive measures to confront food insecurity. Subsequently, the research scrutinizes supply chain optimization to address food security challenges, capitalizing on linear programming and network optimization techniques. These strategies intend to mitigate loss and wastage while streamlining the distribution of agricultural produce from field to fork. In conjunction, the study investigates food distribution networks with a particular focus on network efficiency, accessibility, and equitable food resource allocation. Network analysis tools, complemented by data-driven simulation methodologies, unveil opportunities for augmenting the efficacy of these critical lifelines. This study also considers the ethical implications and privacy concerns associated with the extensive use of data in the realm of food security. The proposed methodology outlines guidelines for responsible data acquisition, storage, and usage. The ultimate aspiration of this research is to forge a nexus between data science and food security policy, bestowing actionable insights to mitigate the ordeal of food insecurity. The holistic approach converging data-driven crop yield forecasts, optimized supply chains, and improved distribution networks aspire to revitalize food security in the most vulnerable regions, elevating the quality of life for millions worldwide.

Keywords: data-driven strategies, crop yield prediction, supply chain optimization, food distribution networks

Procedia PDF Downloads 63

Authors: Arindam Chaudhuri

Abstract:

Classification of data has been actively used for most effective and efficient means of conveying knowledge and information to users. The prima face has always been upon techniques for extracting useful knowledge from data such that returns are maximized. With emergence of huge datasets the existing classification techniques often fail to produce desirable results. The challenge lies in analyzing and understanding characteristics of massive data sets by retrieving useful geometric and statistical patterns. We propose a supervised parallel fuzzy rough support vector machine (PFRSVM) for data classification in cloud environment. The classification is performed by PFRSVM using hyperbolic tangent kernel. The fuzzy rough set model takes care of sensitiveness of noisy samples and handles impreciseness in training samples bringing robustness to results. The membership function is function of center and radius of each class in feature space and is represented with kernel. It plays an important role towards sampling the decision surface. The success of PFRSVM is governed by choosing appropriate parameter values. The training samples are either linear or nonlinear separable. The different input points make unique contributions to decision surface. The algorithm is parallelized with a view to reduce training times. The system is built on support vector machine library using Hadoop implementation of MapReduce. The algorithm is tested on large data sets to check its feasibility and convergence. The performance of classifier is also assessed in terms of number of support vectors. The challenges encountered towards implementing big data classification in machine learning frameworks are also discussed. The experiments are done on the cloud environment available at University of Technology and Management, India. The results are illustrated for Gaussian RBF and Bayesian kernels. The effect of variability in prediction and generalization of PFRSVM is examined with respect to values of parameter C. It effectively resolves outliers’ effects, imbalance and overlapping class problems, normalizes to unseen data and relaxes dependency between features and labels. The average classification accuracy for PFRSVM is better than other classifiers for both Gaussian RBF and Bayesian kernels. The experimental results on both synthetic and real data sets clearly demonstrate the superiority of the proposed technique.

Keywords: FRSVM, Hadoop, MapReduce, PFRSVM

Procedia PDF Downloads 491

Authors: N. C. van der Merwe, J. Oosthuizen, M. F. Makhetha, J. Adams, B. K. Dajee, S-R. Schneider

Abstract:

Women representing high-risk breast cancer families, who tested negative for pathogenic mutations in BRCA1 and BRCA2, are four times more likely to develop breast cancer compared to women in the general population. Sequencing of genes involved in genomic stability and DNA repair led to the identification of novel contributors to familial breast cancer risk. These include BLM and PALB2. Bloom's syndrome is a rare homozygous autosomal recessive chromosomal instability disorder with a high incidence of various types of neoplasia and is associated with breast cancer when in a heterozygous state. PALB2, on the other hand, binds to BRCA2 and together, they partake actively in DNA damage repair. Archived DNA samples of 66 BRCA1/2 negative high-risk breast cancer patients were retrospectively selected based on the presence of an extensive family history of the disease ( > 3 affecteds per family). All coding regions and splice-site boundaries of both genes were screened using High-Resolution Melting Analysis. Samples exhibiting variation were bi-directionally automated Sanger sequenced. The clinical significance of each variant was assessed using various in silico and splice site prediction algorithms. Comprehensive screening identified a total of 11 BLM and 26 PALB2 variants. The variants detected ranged from global to rare and included three novel mutations. Three BLM and two PALB2 likely pathogenic mutations were identified that could account for the disease in these extensive breast cancer families in the absence of BRCA mutations (BLM c.11T > A, p.V4D; BLM c.2603C > T, p.P868L; BLM c.3961G > A, p.V1321I; PALB2 c.421C > T, p.Gln141Ter; PALB2 c.508A > T, p.Arg170Ter). Conclusion: The study confirmed the contribution of pathogenic mutations in BLM and PALB2 to the familial breast cancer burden in South Africa. It explained the presence of the disease in 7.5% of the BRCA1/2 negative families with an extensive family history of breast cancer. Segregation analysis will be performed to confirm the clinical impact of these mutations for each of these families. These results justify the inclusion of both these genes in a comprehensive breast and ovarian next generation sequencing cancer panel and should be screened simultaneously with BRCA1 and BRCA2 as it might explain a significant percentage of familial breast and ovarian cancer in South Africa.

Keywords: Bloom Syndrome, familial breast cancer, PALB2, South Africa

Procedia PDF Downloads 236

Authors: Narendra Akhadkar, Philippe Delcambre

Abstract:

In power protection and control products, assembly process variations due to the individual parts manufactured from single or multi-cavity tooling is a major problem. The dimensional and geometrical variations on the individual parts, in the form of manufacturing tolerances and assembly tolerances, are sources of clearance in the kinematic joints, polarization effect in the joints, and tolerance stack-up. All these variations adversely affect the quality of product, functionality, cost, and time-to-market. Variation simulation analysis may be used in the early product design stage to predict such uncertainties. Usually, variations exist in both manufacturing processes and materials. In the tolerance analysis, the effect of the dimensional and geometrical variations of the individual parts on the functional characteristics (conditions) of the final assembled products are studied. A functional characteristic of the product may be affected by a set of interrelated dimensions (functional parameters) that usually form a geometrical closure in a 3D chain. In power protection and control products, the prerequisite is: when a fault occurs in the electrical network, the product must respond quickly to react and break the circuit to clear the fault. Usually, the response time is in milliseconds. Any failure in clearing the fault may result in severe damage to the equipment or network, and human safety is at stake. In this article, we have investigated two important functional characteristics that are associated with the robust performance of the product. It is demonstrated that the experimental data obtained at the Schneider Electric Laboratory prove the very good prediction capabilities of the variation simulation performed using CETOL (tolerance analysis software) in an industrial context. Especially, this study allows design engineers to better understand the critical parts in the product that needs to be manufactured with good, capable tolerances. On the contrary, some parts are not critical for the functional characteristics (conditions) of the product and may lead to some reduction of the manufacturing cost, ensuring robust performance. The capable tolerancing is one of the most important aspects in product and manufacturing process design. In the case of miniature circuit breaker (MCB), the product's quality and its robustness are mainly impacted by two aspects: (1) allocation of design tolerances between the components of a mechanical assembly and (2) manufacturing tolerances in the intermediate machining steps of component fabrication.

Keywords: geometrical variation, product robustness, tolerance analysis, variation simulation

Procedia PDF Downloads 164

Authors: Hanwei Gao, Louis Jezequel, Eric Cabrol, Bernard Vitry

Abstract:

The chassis system is composed of complex elements that take up all the loads from the tire-ground contact area and thus it plays an important role in numerous specifications such as durability, comfort, crash, etc. During the development of new vehicle projects in Renault, durability validation is always the main focus while deployment of comfort comes later in the project. Therefore, sometimes design choices have to be reconsidered because of the natural incompatibility between these two specifications. Besides, robustness is also an important point of concern as it is related to manufacturing costs as well as the performance after the ageing of components like shock absorbers. In this paper an approach is proposed aiming to realize a multi-objective optimization between chassis endurance and comfort while taking the random factors into consideration. The adaptive-sparse polynomial chaos expansion method (PCE) with Chebyshev polynomial series has been applied to predict responses’ uncertainty intervals of a system according to its uncertain-but-bounded parameters. The approach can be divided into three steps. First an initial design of experiments is realized to build the response surfaces which represent statistically a black-box system. Secondly within several iterations an optimum set is proposed and validated which will form a Pareto front. At the same time the robustness of each response, served as additional objectives, is calculated from the pre-defined parameter intervals and the response surfaces obtained in the first step. Finally an inverse strategy is carried out to determine the parameters’ tolerance combination with a maximally acceptable degradation of the responses in terms of manufacturing costs. A quarter car model has been tested as an example by applying the road excitations from the actual road measurements for both endurance and comfort calculations. One indicator based on the Basquin’s law is defined to compare the global chassis durability of different parameter settings. Another indicator related to comfort is obtained from the vertical acceleration of the sprung mass. An optimum set with best robustness has been finally obtained and the reference tests prove a good robustness prediction of Chebyshev PCE method. This example demonstrates the effectiveness and reliability of the approach, in particular its ability to save computational costs for a complex system.

Keywords: chassis durability, Chebyshev polynomials, multi-objective optimization, polynomial chaos expansion, ride comfort, robust design

Procedia PDF Downloads 153

Authors: Mustafa Alhamdi

Abstract:

Industrial application to classify gamma rays and neutron events is investigated in this study using deep machine learning. The identification using a convolutional neural network and recursive neural network showed a significant improvement in predication accuracy in a variety of applications. The ability to identify the isotope type and activity from spectral information depends on feature extraction methods, followed by classification. The features extracted from the spectrum profiles try to find patterns and relationships to present the actual spectrum energy in low dimensional space. Increasing the level of separation between classes in feature space improves the possibility to enhance classification accuracy. The nonlinear nature to extract features by neural network contains a variety of transformation and mathematical optimization, while principal component analysis depends on linear transformations to extract features and subsequently improve the classification accuracy. In this paper, the isotope spectrum information has been preprocessed by finding the frequencies components relative to time and using them as a training dataset. Fourier transform implementation to extract frequencies component has been optimized by a suitable windowing function. Training and validation samples of different isotope profiles interacted with CdTe crystal have been simulated using Geant4. The readout electronic noise has been simulated by optimizing the mean and variance of normal distribution. Ensemble learning by combing voting of many models managed to improve the classification accuracy of neural networks. The ability to discriminate gamma and neutron events in a single predication approach using deep machine learning has shown high accuracy using deep learning. The paper findings show the ability to improve the classification accuracy by applying the spectrogram preprocessing stage to the gamma and neutron spectrums of different isotopes. Tuning deep machine learning models by hyperparameter optimization of neural network models enhanced the separation in the latent space and provided the ability to extend the number of detected isotopes in the training database. Ensemble learning contributed significantly to improve the final prediction.

Keywords: machine learning, nuclear physics, Monte Carlo simulation, noise estimation, feature extraction, classification

Procedia PDF Downloads 151

Authors: E. Menga, S. Hernandez

Abstract:

Finite Element Models (FEMs) are widely used in order to study and predict the dynamic properties of structures and usually, the prediction can be obtained with much more accuracy in the case of a single component than in the case of assemblies. Especially for structural dynamics studies, in the low and middle frequency range, most complex FEMs can be seen as assemblies made by linear components joined together at interfaces. From a modelling and computational point of view, these types of joints can be seen as localized sources of stiffness and damping and can be modelled as lumped spring/damper elements, most of time, characterized by nonlinear constitutive laws. On the other side, most of FE programs are able to run nonlinear analysis in time-domain. They treat the whole structure as nonlinear, even if there is one nonlinear degree of freedom (DOF) out of thousands of linear ones, making the analysis unnecessarily expensive from a computational point of view. In this work, a methodology in order to obtain the nonlinear frequency response of structures, whose nonlinearities can be considered as localized sources, is presented. The work extends the well-known Structural Dynamic Modification Method (SDMM) to a nonlinear set of modifications, and allows getting the Nonlinear Frequency Response Functions (NLFRFs), through an ‘updating’ process of the Linear Frequency Response Functions (LFRFs). A brief summary of the analytical concepts is given, starting from the linear formulation and understanding what the implications of the nonlinear one, are. The response of the system is formulated in both: time and frequency domain. First the Modal Database is extracted and the linear response is calculated. Secondly the nonlinear response is obtained thru the NL SDMM, by updating the underlying linear behavior of the system. The methodology, implemented in MATLAB, has been successfully applied to estimate the nonlinear frequency response of two systems. The first one is a two DOFs spring-mass-damper system, and the second example takes into account a full aircraft FE Model. In spite of the different levels of complexity, both examples show the reliability and effectiveness of the method. The results highlight a feasible and robust procedure, which allows a quick estimation of the effect of localized nonlinearities on the dynamic behavior. The method is particularly powerful when most of the FE Model can be considered as acting linearly and the nonlinear behavior is restricted to few degrees of freedom. The procedure is very attractive from a computational point of view because the FEM needs to be run just once, which allows faster nonlinear sensitivity analysis and easier implementation of optimization procedures for the calibration of nonlinear models.

Keywords: frequency response, nonlinear dynamics, structural dynamic modification, softening effect, rubber

Procedia PDF Downloads 266

Authors: Markus Rütten, Olaf Wünsch

Abstract:

Non-Newtonian fluid properties can change the flow behaviour significantly, its prediction is more difficult when thermal effects come into play. Hence, the focal point of this work is the wake flow behind a heated circular cylinder in the laminar vortex shedding regime for thermo-viscous shear thinning fluids. In the case of isothermal flows of Newtonian fluids the vortex shedding regime is characterised by a distinct Reynolds number and an associated Strouhal number. In the case of thermo-viscous shear thinning fluids the flow regime can significantly change in dependence of the temperature of the viscous wall of the cylinder. The Reynolds number alters locally and, consequentially, the Strouhal number globally. In the present CFD study the temperature dependence of the Reynolds and Strouhal number is investigated for the flow of a Carreau fluid around a heated cylinder. The temperature dependence of the fluid viscosity has been modelled by applying the standard Williams-Landel-Ferry (WLF) equation. In the present simulation campaign thermal boundary conditions have been varied over a wide range in order to derive a relation between dimensionless heat transfer, Reynolds and Strouhal number. Together with the shear thinning due to the high shear rates close to the cylinder wall this leads to a significant decrease of viscosity of three orders of magnitude in the nearfield of the cylinder and a reduction of two orders of magnitude in the wake field. Yet the shear thinning effect is able to change the flow topology: a complex K´arm´an vortex street occurs, also revealing distinct characteristic frequencies associated with the dominant and sub-dominant vortices. Heating up the cylinder wall leads to a delayed flow separation and narrower wake flow, giving lesser space for the sequence of counter-rotating vortices. This spatial limitation does not only reduce the amplitude of the oscillating wake flow it also shifts the dominant frequency to higher frequencies, furthermore it damps higher harmonics. Eventually the locally heated wake flow smears out. Eventually, the CFD simulation results of the systematically varied thermal flow parameter study have been used to describe a relation for the main characteristic order parameters.

Keywords: heat transfer, thermo-viscous fluids, shear thinning, vortex shedding

Procedia PDF Downloads 298

Authors: Taylan Kabbani, Ekrem Duman

Abstract:

The design of adaptive systems that take advantage of financial markets while reducing the risk can bring more stagnant wealth into the global market. However, most efforts made to generate successful deals in trading financial assets rely on Supervised Learning (SL), which suffered from various limitations. Deep Reinforcement Learning (DRL) offers to solve these drawbacks of SL approaches by combining the financial assets price "prediction" step and the "allocation" step of the portfolio in one unified process to produce fully autonomous systems capable of interacting with its environment to make optimal decisions through trial and error. In this paper, a continuous action space approach is adopted to give the trading agent the ability to gradually adjust the portfolio's positions with each time step (dynamically re-allocate investments), resulting in better agent-environment interaction and faster convergence of the learning process. In addition, the approach supports the managing of a portfolio with several assets instead of a single one. This work represents a novel DRL model to generate profitable trades in the stock market, effectively overcoming the limitations of supervised learning approaches. We formulate the trading problem, or what is referred to as The Agent Environment as Partially observed Markov Decision Process (POMDP) model, considering the constraints imposed by the stock market, such as liquidity and transaction costs. More specifically, we design an environment that simulates the real-world trading process by augmenting the state representation with ten different technical indicators and sentiment analysis of news articles for each stock. We then solve the formulated POMDP problem using the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm, which can learn policies in high-dimensional and continuous action spaces like those typically found in the stock market environment. From the point of view of stock market forecasting and the intelligent decision-making mechanism, this paper demonstrates the superiority of deep reinforcement learning in financial markets over other types of machine learning such as supervised learning and proves its credibility and advantages of strategic decision-making.

Keywords: the stock market, deep reinforcement learning, MDP, twin delayed deep deterministic policy gradient, sentiment analysis, technical indicators, autonomous agent

Procedia PDF Downloads 178

Authors: Ruzanna Muradyan

Abstract:

Medical data mining is a crucial field in contemporary healthcare that offers cutting-edge tactics with enormous potential to transform patient care. This abstract examines how sophisticated data mining techniques could transform the healthcare industry, with a special focus on how they might improve patient outcomes. Healthcare data repositories have dynamically evolved, producing a rich tapestry of different, multi-dimensional information that includes genetic profiles, lifestyle markers, electronic health records, and more. By utilizing data mining techniques inside this vast library, a variety of prospects for precision medicine, predictive analytics, and insight production become visible. Predictive modeling for illness prediction, risk stratification, and therapy efficacy evaluations are important points of focus. Healthcare providers may use this abundance of data to tailor treatment plans, identify high-risk patient populations, and forecast disease trajectories by applying machine learning algorithms and predictive analytics. Better patient outcomes, more efficient use of resources, and early treatments are made possible by this proactive strategy. Furthermore, data mining techniques act as catalysts to reveal complex relationships between apparently unrelated data pieces, providing enhanced insights into the cause of disease, genetic susceptibilities, and environmental factors. Healthcare practitioners can get practical insights that guide disease prevention, customized patient counseling, and focused therapies by analyzing these associations. The abstract explores the problems and ethical issues that come with using data mining techniques in the healthcare industry. In order to properly use these approaches, it is essential to find a balance between data privacy, security issues, and the interpretability of complex models. Finally, this abstract demonstrates the revolutionary power of modern data mining methodologies in transforming the healthcare sector. Healthcare practitioners and researchers can uncover unique insights, enhance clinical decision-making, and ultimately elevate patient care to unprecedented levels of precision and efficacy by employing cutting-edge methodologies.

Keywords: data mining, healthcare, patient care, predictive analytics, precision medicine, electronic health records, machine learning, predictive modeling, disease prognosis, risk stratification, treatment efficacy, genetic profiles, precision health

Procedia PDF Downloads 63

Authors: Ella Tyuryumina, Alexey Neznanov

Abstract:

This study is an attempt to obtain reliable data on the natural history of breast cancer growth. We analyze the opportunities for using classical mathematical models (exponential and logistic tumor growth models, Gompertz and von Bertalanffy tumor growth models) to try to describe growth of the primary tumor and the secondary distant metastases of human breast cancer. The research aim is to improve predicting accuracy of breast cancer progression using an original mathematical model referred to CoMPaS and corresponding software. We are interested in: 1) modelling the whole natural history of the primary tumor and the secondary distant metastases; 2) developing adequate and precise CoMPaS which reflects relations between the primary tumor and the secondary distant metastases; 3) analyzing the CoMPaS scope of application; 4) implementing the model as a software tool. The foundation of the CoMPaS is the exponential tumor growth model, which is described by determinate nonlinear and linear equations. The CoMPaS corresponds to TNM classification. It allows to calculate different growth periods of the primary tumor and the secondary distant metastases: 1) ‘non-visible period’ for the primary tumor; 2) ‘non-visible period’ for the secondary distant metastases; 3) ‘visible period’ for the secondary distant metastases. The CoMPaS is validated on clinical data of 10-years and 15-years survival depending on the tumor stage and diameter of the primary tumor. The new predictive tool: 1) is a solid foundation to develop future studies of breast cancer growth models; 2) does not require any expensive diagnostic tests; 3) is the first predictor which makes forecast using only current patient data, the others are based on the additional statistical data. The CoMPaS model and predictive software: a) fit to clinical trials data; b) detect different growth periods of the primary tumor and the secondary distant metastases; c) make forecast of the period of the secondary distant metastases appearance; d) have higher average prediction accuracy than the other tools; e) can improve forecasts on survival of breast cancer and facilitate optimization of diagnostic tests. The following are calculated by CoMPaS: the number of doublings for ‘non-visible’ and ‘visible’ growth period of the secondary distant metastases; tumor volume doubling time (days) for ‘non-visible’ and ‘visible’ growth period of the secondary distant metastases. The CoMPaS enables, for the first time, to predict ‘whole natural history’ of the primary tumor and the secondary distant metastases growth on each stage (pT1, pT2, pT3, pT4) relying only on the primary tumor sizes. Summarizing: a) CoMPaS describes correctly the primary tumor growth of IA, IIA, IIB, IIIB (T1-4N0M0) stages without metastases in lymph nodes (N0); b) facilitates the understanding of the appearance period and inception of the secondary distant metastases.

Keywords: breast cancer, exponential growth model, mathematical model, metastases in lymph nodes, primary tumor, survival

Procedia PDF Downloads 341

Authors: Md. Fazlul Karim, Ashik Sharfaraz, Aysha Ferdoushi

Abstract:

Background: Computational medicinal chemistry approaches are used for designing and identifying new drug-like molecules, predicting properties and pharmacological activities, and optimizing lead compounds in drug development. SIRT7, a nicotinamide adenine dinucleotide (NAD+)-dependent deacylase which regulates aging, is an emerging target for cancer therapy with mounting evidence that SIRT7 downregulation plays important roles in reversing cancer phenotypes and suppressing tumor growth. Activation or altered expression of SIRT7 is associated with the progression and invasion of various cancers, including liver, breast, gastric, prostate, and non-small cell lung cancer. Objectives: The goal of this work was to identify potent and selective bioactive candidate inhibitors of SIRT7 by in silico screening of small molecule compounds obtained from Nigella sativa (N. sativa). Methods: SIRT7 structure was retrieved from The Research Collaboratory for Structural Bioinformatics Protein Data Bank (RCSB PDB), and its active site was identified using CASTp and metaPocket. Molecular docking simulation was performed with PyRx 0.8 virtual screening software. Drug-likeness properties were tested using SwissADME and pkCSM. In silico toxicity was evaluated by Osiris Property Explorer. Bioactivity was predicted by Molinspiration software. Antitumor activity was screened for Prediction of Activity Spectra for Substances (PASS) using Way2Drug web server. Molecular dynamics (MD) simulation was carried out by Desmond v3.6 package. Results: A total of 159 bioactive compounds from the N. Sativa were screened against the SIRT7 enzyme. Five bioactive compounds: chrysin (CID:5281607), pinocembrin (CID:68071), nigellidine (CID:136828302), nigellicine (CID:11402337), and epicatechin (CID:72276) were identified as potent SIRT7 anti-cancer candidates after docking score evaluation and applying Lipinski's Rule of Five. Finally, MD simulation identified Chrysin as the top SIRT7 anti-cancer candidate molecule. Conclusion: Chrysin, which shows a potential inhibitory effect against SIRT7, can act as a possible anti-cancer drug candidate. This inhibitor warrants further evaluation to check its pharmacokinetics and pharmacodynamics properties both in vitro and in vivo.

Keywords: SIRT7, antitumor, molecular docking, molecular dynamics simulation

Procedia PDF Downloads 80

Authors: Christian Ihwiwhu, Nnamdi Obioha, Udeme John, Edward Bobade, Oghenerunor Bekibele, Adedeji Awujoola, Ibi-Ada Itotoi

Abstract:

Understanding the complexity in the distribution of hydrocarbon in a simple structure with flow baffles and connectivity issues is critical in targeting and developing the remaining pay in a mature asset. Subtle facies changes (heterogeneity) can have a drastic impact on reservoir fluids movement, and this can be crucial to identifying sweet spots in mature fields. This study aims to evaluate selected reservoirs in Ovhor Field, Niger Delta, Nigeria, with the objective of optimising production from the field by targeting undeveloped oil reserves, bypassed pay, and gaining an improved understanding of the selected reservoirs to increase the company’s reservoir limits. The task at the Ovhor field is complicated by poor stratigraphic seismic resolution over the field. 3-D geological (sedimentology and stratigraphy) interpretation, use of results from quantitative interpretation, and proper understanding of production data have been used in recognizing flow baffles and undeveloped compartments in the field. The full field 3-D model has been constructed in such a way as to capture heterogeneities and the various compartments in the field to aid the proper simulation of fluid flow in the field for future production prediction, proper history matching and design of good trajectories to adequately target undeveloped oil in the field. Reservoir property models (porosity, permeability, and net-to-gross) have been constructed by biasing log interpreted properties to a defined environment of deposition model whose interpretation captures the heterogeneities expected in the studied reservoirs. At least, two scenarios have been modelled for most of the studied reservoirs to capture the range of uncertainties we are dealing with. The total original oil in-place volume for the four reservoirs studied is 157 MMstb. The cumulative oil and gas production from the selected reservoirs are 67.64 MMstb and 9.76 Bscf respectively, with current production rate of about 7035 bopd and 4.38 MMscf/d (as at 31/08/2019). Dynamic simulation and production forecast on the 4 reservoirs gave an undeveloped reserve of about 3.82 MMstb from two (2) identified oil restoration activities. These activities include side-tracking and re-perforation of existing wells. This integrated approach led to the identification of bypassed oil in some areas of the selected reservoirs and an improved understanding of the studied reservoirs. New wells have/are being drilled now to test the results of our studies, and the results are very confirmatory and satisfying.

Keywords: facies, flow baffle, bypassed pay, heterogeneities, history matching, reservoir limit

Procedia PDF Downloads 129

Authors: Taikchan Lildar, Ayesha Samad, Suraj Sookhu

Abstract:

Atrial fibrillation with rapid ventricular response results in the loss of atrial kick and shortened ventricular filling time, which often leads to decompensated heart failure. Pharmacologic rhythm control is the treatment of choice, and patients frequently benefit from the restoration of sinus rhythm. When pharmacologic treatment is unsuccessful or a patient declines hemodynamically, direct cardioversion is the treatment of choice. Torsades de pointes or “twisting of the points'' in French, is a rare but under-appreciated risk of cardioversion therapy and accounts for a significant number of sudden cardiac death each year. A 61-year-old female with no significant past medical history presented to the Emergency Department with worsening dyspnea. An electrocardiogram showed atrial fibrillation with rapid ventricular response, and a chest X-ray was significant for bilateral pulmonary vascular congestion. Full-dose anticoagulation and diuresis were initiated with moderate improvement in symptoms. A transthoracic echocardiogram revealed biventricular systolic dysfunction with a left ventricular ejection fraction of 30%. After consultation with an electrophysiologist, the consensus was to proceed with the restoration of sinus rhythm, which would likely improve the patient’s heart failure symptoms and possibly the ejection fraction. A transesophageal echocardiogram was negative for left atrial appendage thrombus; the patient was treated with a loading dose of amiodarone and underwent successful direct current cardioversion with 200 Joules. The patient was placed on telemetry monitoring for 24 hours and was noted to have frequent premature ventricular contractions with subsequent degeneration to torsades de pointes. The patient was found unresponsive and pulseless; cardiopulmonary resuscitation was initiated with cardioversion, and return of spontaneous circulation was achieved after four minutes to normal sinus rhythm. Post-cardiac arrest electrocardiogram showed sinus bradycardia with heart-rate corrected QT interval of 592 milliseconds. The patient continued to have frequent premature ventricular contractions and required two additional cardioversions to achieve a return of spontaneous circulation with intravenous magnesium and lidocaine. An automatic implantable cardioverter-defibrillator was subsequently implanted for secondary prevention of sudden cardiac death. The backup pacing rate of the automatic implantable cardioverter-defibrillator was set higher than usual in an attempt to prevent premature ventricular contractions-induced torsades de pointes. The patient did not have any further ventricular arrhythmias after implantation of the automatic implantable cardioverter-defibrillator. Overdrive pacing is a method utilized to treat premature ventricular contractions-induced torsades de pointes by preventing a patient’s susceptibility to R on T-wave-induced ventricular arrhythmias. Pacing at a rate of 90 beats per minute succeeded in controlling the arrhythmia without the need for traumatic cardiac defibrillation. In our patient, conversion of atrial fibrillation with rapid ventricular response to normal sinus rhythm resulted in a slower heart rate and an increased probability of premature ventricular contraction occurring on the T-wave and ensuing ventricular arrhythmia. This case highlights direct current cardioversion for atrial fibrillation with rapid ventricular response resulting in persistent ventricular arrhythmia requiring an automatic implantable cardioverter-defibrillator placement with overdrive pacing to prevent a recurrence.

Keywords: refractory atrial fibrillation, atrial fibrillation, overdrive pacing, torsades de pointes

Procedia PDF Downloads 149

Authors: Sandra Smart-Akande, Chaminda Hewage, Imtiaz Khan, Thanuja Mallikarachchi

Abstract:

According to UK police-recorded data, there has been a substantial increase in knife-related crime and youth violence in the UK since 2014 particularly in the London boroughs. These crime rates are disproportionally distributed across London with the majority of these crimes occurring in the highly deprived areas of London and among young people aged 11 to 24 with large discrepancies across ethnicity, age, gender and borough of residence. Comprehensive studies and literature have identified risk factors associated with a knife carrying among youth to be Adverse Childhood Experience (ACEs), poor mental health, school or social exclusion, drug dealing, drug using, victim of violent crime, bullying, peer pressure or gang involvement, just to mention a few. ACEs are potentially traumatic events that occur in childhood, this can be experiences or stressful events in the early life of a child and can lead to an increased risk of damaging health or social outcomes in the latter life of the individual. Research has shown that children or youths involved in youth violence have had childhood experience characterised by disproportionate adverse childhood experiences and substantial literature link ACEs to be associated with criminal or delinquent behavior. ACEs are commonly grouped by researchers into: Abuse (Physical, Verbal, Sexual), Neglect (Physical, Emotional) and Household adversities (Mental Illness, Incarcerated relative, Domestic violence, Parental Separation or Bereavement). To the author's best knowledge, no study to date has investigated how household mental health (mental health of a parent or mental health of a child) and domestic violence (domestic violence on a parent or domestic violence on a child) is related to knife homicides across the local authorities areas of London. This study seeks to address the gap by examining a large sample of data from the London Metropolitan Police Force and Characteristics of Children in Need data from the UK Department for Education. The aim of this review is to identify and synthesise evidence from data and a range of literature to identify the relationship between adverse childhood experiences and youth violence in the UK. Understanding the link between ACEs and future outcomes can support preventative action.

Keywords: adverse childhood experiences, domestic violence, mental health, youth violence, prediction analysis, London knife crime

Procedia PDF Downloads 121

Authors: Claudia García-García, Catalina B. García-García, Román Salmerón-Gómez

Abstract:

Much economic-environmental research includes the analysis of possible interactions by using Moderated Regression Analysis (MRA), which is a specific application of multiple linear regression analysis. This methodology allows analyzing how the effect of one of the independent variables is moderated by a second independent variable by adding a cross-product term between them as an additional explanatory variable. Due to the very specification of the methodology, the moderated factor is often highly correlated with the constitutive terms. Thus, great multicollinearity problems arise. The appearance of strong multicollinearity in a model has important consequences. Inflated variances of the estimators may appear, there is a tendency to consider non-significant regressors that they probably are together with a very high coefficient of determination, incorrect signs of our coefficients may appear and also the high sensibility of the results to small changes in the dataset. Finally, the high relationship among explanatory variables implies difficulties in fixing the individual effects of each one on the model under study. These consequences shifted to the moderated analysis may imply that it is not worth including an interaction term that may be distorting the model. Thus, it is important to manage the problem with some methodology that allows for obtaining reliable results. After a review of those works that applied the MRA among the ten top journals of the field, it is clear that multicollinearity is mostly disregarded. Less than 15% of the reviewed works take into account potential multicollinearity problems. To overcome the issue, this work studies the possible application of recent methodologies to MRA. Particularly, the raised regression is analyzed. This methodology mitigates collinearity from a geometrical point of view: the collinearity problem arises because the variables under study are very close geometrically, so by separating both variables, the problem can be mitigated. Raise regression maintains the available information and modifies the problematic variables instead of deleting variables, for example. Furthermore, the global characteristics of the initial model are also maintained (sum of squared residuals, estimated variance, coefficient of determination, global significance test and prediction). The proposal is implemented to data from countries of the European Union during the last year available regarding greenhouse gas emissions, per capita GDP and a dummy variable that represents the topography of the country. The use of a dummy variable as the moderator is a special variant of MRA, sometimes called “subgroup regression analysis.” The main conclusion of this work is that applying new techniques to the field can improve in a substantial way the results of the analysis. Particularly, the use of raised regression mitigates great multicollinearity problems, so the researcher is able to rely on the interaction term when interpreting the results of a particular study.

Keywords: multicollinearity, MRA, interaction, raise

Procedia PDF Downloads 107

Authors: Virendra Nath, Vipin Kumar

Abstract:

Background: TypeII Diabetes mellitus is a foremost health problem worldwide, predisposing to increased mortality and morbidity. Undesirable effects of the current medications have prompted the researcher to develop more potential drug(s) against the disease. The peroxisome proliferator-activated receptors (PPARs) are members of the nuclear receptors family and take part in a vital role in the regulation of metabolic equilibrium. They can induce or repress genes associated with adipogenesis, lipid, and glucose metabolism. Aims: Investigation of PPARα/γ agonistic hits were screened by hierarchical virtual screening followed by molecular dynamics simulation and knowledge-based structure-activity relation (SAR) analysis using approved PPAR α/γ dual agonist. Methods: The PPARα/γ agonistic activity of compounds was searched by using Maestro through structure-based virtual screening and molecular dynamics (MD) simulation application. Virtual screening of nuclear-receptor ligands was done, and the binding modes with protein-ligand interactions of newer entity(s) were investigated. Further, binding energy prediction, Stability studies using molecular dynamics (MD) simulation of PPARα and γ complex was performed with the most promising hit along with the structural comparative analysis of approved PPARα/γ agonists with screened hit was done for knowledge-based SAR. Results and Discussion: The silicone chip-based approach recognized the most capable nine hits and had better predictive binding energy as compared to the reference drug compound (Tesaglitazar). In this study, the key amino acid residues of binding pockets of both targets PPARα/γ were acknowledged as essential and were found to be associated in the key interactions with the most potential dual hit (ChemDiv-3269-0443). Stability studies using molecular dynamics (MD) simulation of PPARα and γ complex was performed with the most promising hit and found root mean square deviation (RMSD) stabile around 2Å and 2.1Å, respectively. Frequency distribution data also revealed that the key residues of both proteins showed maximum contacts with a potent hit during the MD simulation of 20 nanoseconds (ns). The knowledge-based SAR studies of PPARα/γ agonists were studied using 2D structures of approved drugs like aleglitazar, tesaglitazar, etc. for successful designing and synthesis of compounds PPARγ agonistic candidates with anti-hyperlipidimic potential.

Keywords: computational, diabetes, PPAR, simulation

Procedia PDF Downloads 103

Authors: Mohd Khairul Bazli Mohd Aziz, Fadhilah Yusof, Zulkifli Yusop, Zalina Mohd Daud, Mohammad Afif Kasno

Abstract:

Recent developed rainfall network design techniques have been discussed and compared by many researchers worldwide due to the demand of acquiring higher levels of accuracy from collected data. In many studies, rain-gauge networks are designed to provide good estimation for areal rainfall and for flood modelling and prediction. In a certain study, even using lumped models for flood forecasting, a proper gauge network can significantly improve the results. Therefore existing rainfall network in Johor must be optimized and redesigned in order to meet the required level of accuracy preset by rainfall data users. The well-known geostatistics method (variance-reduction method) that is combined with simulated annealing was used as an algorithm of optimization in this study to obtain the optimal number and locations of the rain gauges. Rain gauge network structure is not only dependent on the station density; station location also plays an important role in determining whether information is acquired accurately. The existing network of 84 rain gauges in Johor is optimized and redesigned by using rainfall, humidity, solar radiation, temperature and wind speed data during monsoon season (November – February) for the period of 1975 – 2008. Three different semivariogram models which are Spherical, Gaussian and Exponential were used and their performances were also compared in this study. Cross validation technique was applied to compute the errors and the result showed that exponential model is the best semivariogram. It was found that the proposed method was satisfied by a network of 64 rain gauges with the minimum estimated variance and 20 of the existing ones were removed and relocated. An existing network may consist of redundant stations that may make little or no contribution to the network performance for providing quality data. Therefore, two different cases were considered in this study. The first case considered the removed stations that were optimally relocated into new locations to investigate their influence in the calculated estimated variance and the second case explored the possibility to relocate all 84 existing stations into new locations to determine the optimal position. The relocations of the stations in both cases have shown that the new optimal locations have managed to reduce the estimated variance and it has proven that locations played an important role in determining the optimal network.

Keywords: geostatistics, simulated annealing, semivariogram, optimization

Procedia PDF Downloads 304

Authors: Zhihui Liu, Dongmei Hao, Qian Qiu, Yang An, Lin Yang, Song Zhang, Yimin Yang, Xuwen Li, Dingchang Zheng

Abstract:

Premature birth, defined as birth before 37 completed weeks of gestation is a leading cause of neonatal morbidity and mortality and has long-term adverse consequences for health. It has recently been reported that the worldwide preterm birth rate is around 10%. The existing measurement techniques for diagnosing preterm delivery include tocodynamometer, ultrasound and fetal fibronectin. However, they are subjective, or suffer from high measurement variability and inaccurate diagnosis and prediction of preterm labor. Electrohysterography (EHG) method based on recording of uterine electrical activity by electrodes attached to maternal abdomen, is a promising method to assess uterine activity and diagnose preterm labor. The purpose of this study is to analyze the difference of EHG signal features between term labor and preterm labor. Free access database was used with 300 signals acquired in two groups of pregnant women who delivered at term (262 cases) and preterm (38 cases). Among them, EHG signals from 38 term labor and 38 preterm labor were preprocessed with band-pass Butterworth filters of 0.08–4Hz. Then, EHG signal features were extracted, which comprised classical time domain description including root mean square and zero-crossing number, spectral parameters including peak frequency, mean frequency and median frequency, wavelet packet coefficients, autoregression (AR) model coefficients, and nonlinear measures including maximal Lyapunov exponent, sample entropy and correlation dimension. Their statistical significance for recognition of two groups of recordings was provided. The results showed that mean frequency of preterm labor was significantly smaller than term labor (p < 0.05). 5 coefficients of AR model showed significant difference between term labor and preterm labor. The maximal Lyapunov exponent of early preterm (time of recording < the 26th week of gestation) was significantly smaller than early term. The sample entropy of late preterm (time of recording > the 26th week of gestation) was significantly smaller than late term. There was no significant difference for other features between the term labor and preterm labor groups. Any future work regarding classification should therefore focus on using multiple techniques, with the mean frequency, AR coefficients, maximal Lyapunov exponent and the sample entropy being among the prime candidates. Even if these methods are not yet useful for clinical practice, they do bring the most promising indicators for the preterm labor.

Keywords: electrohysterogram, feature, preterm labor, term labor

Procedia PDF Downloads 572

Authors: Dangut Maren David, Skaf Zakwan

Abstract:

Adequate monitoring of vehicle component in other to obtain high uptime is the goal of predictive maintenance, the major challenge faced by businesses in industries is the significant cost associated with a delay in service delivery due to system downtime. Most of those businesses are interested in predicting those problems and proactively prevent them in advance before it occurs, which is the core advantage of Prognostic Health Management (PHM) application. The recent emergence of industry 4.0 or industrial internet of things (IIoT) has led to the need for monitoring systems activities and enhancing system-to-system or component-to- component interactions, this has resulted to a large generation of data known as big data. Analysis of big data represents an increasingly important, however, due to complexity inherently in the dataset such as imbalance classification problems, it becomes extremely difficult to build a model with accurate high precision. Data-driven predictive modeling for condition-based maintenance (CBM) has recently drowned research interest with growing attention to both academics and industries. The large data generated from industrial process inherently comes with a different degree of complexity which posed a challenge for analytics. Thus, imbalance classification problem exists perversely in industrial datasets which can affect the performance of learning algorithms yielding to poor classifier accuracy in model development. Misclassification of faults can result in unplanned breakdown leading economic loss. In this paper, an advanced approach for handling imbalance classification problem is proposed and then a prognostic model for predicting aircraft component replacement is developed to predict component replacement in advanced by exploring aircraft historical data, the approached is based on hybrid ensemble-based method which improves the prediction of the minority class during learning, we also investigate the impact of our approach on multiclass imbalance problem. We validate the feasibility and effectiveness in terms of the performance of our approach using real-world aircraft operation and maintenance datasets, which spans over 7 years. Our approach shows better performance compared to other similar approaches. We also validate our approach strength for handling multiclass imbalanced dataset, our results also show good performance compared to other based classifiers.

Keywords: prognostics, data-driven, imbalance classification, deep learning

Procedia PDF Downloads 175

Authors: Mindaye Teshome, Evaldo Muñoz Braz, Carlos M. M. Eleto Torres, Patricia Mattos

Abstract:

Sustainable forest management planning requires up-to-date information on the structure, standing volume, biomass, and growth rate of trees from a given forest. This kind of information is lacking in many forests in Ethiopia. The objective of this study was to quantify the population structure, diameter growth rate, and standing volume of wood from Juniperus procera trees in the Chilimo forest. A total of 163 sample plots were set up in the forest to collect the relevant vegetation data. Growth ring measurements were conducted on stem disc samples collected from 12 J. procera trees. Diameter and height measurements were recorded from a total of 1399 individual trees with dbh ≥ 2 cm. The growth rate, maximum current and mean annual increments, minimum logging diameter, and cutting cycle were estimated, and alternative cutting cycles were established. Using these data, the harvestable volume of wood was projected by alternating four minimum logging diameters and five cutting cycles following the stand table projection method. The results show that J. procera trees have an average density of 183 stems ha⁻¹, a total basal area of 12.1 m² ha⁻¹, and a standing volume of 98.9 m³ ha⁻¹. The mean annual diameter growth ranges between 0.50 and 0.65 cm year⁻¹ with an overall mean of 0.59 cm year⁻¹. The population of J. procera tree followed a reverse J-shape diameter distribution pattern. The maximum current annual increment in volume (CAI) occurred at around 49 years when trees reached 30 cm in diameter. Trees showed the maximum mean annual increment in volume (MAI) around 91 years, with a diameter size of 50 cm. The simulation analysis revealed that 40 cm MLD and a 15-year cutting cycle are the best minimum logging diameter and cutting cycle. This combination showed the largest harvestable volume of wood potential, volume increments, and a 35% recovery of the initially harvested volume. It is concluded that the forest is well stocked and has a large amount of harvestable volume of wood from J. procera trees. This will enable the country to partly meet the national wood demand through domestic wood production. The use of the current population structure and diameter growth data from tree ring analysis enables the exact prediction of the harvestable volume of wood. The developed model supplied an idea about the productivity of the J. procera tree population and enables policymakers to develop specific management criteria for wood harvesting.

Keywords: logging, growth model, cutting cycle, minimum logging diameter

Procedia PDF Downloads 89

Authors: Lahcen Daoudi, Fatima Zahra Omdi, Abldelali Gourfi

Abstract:

In Morocco, as in most Mediterranean countries, water scarcity is a common situation because of low and unevenly distributed rainfall. The expansions of irrigated lands, as well as the growth of urban and industrial areas and tourist resorts, contribute to an increase of water demand. Therefore in the 1960s Morocco embarked on an ambitious program to increase the number of dams to boost water retention capacity. However, the decrease in the capacity of these reservoirs caused by sedimentation is a major problem; it is estimated at 75 million m3/year. Dams and reservoirs became unusable for their intended purposes due to sedimentation in large rivers that result from soil erosion. Soil erosion presents an important driving force in the process affecting the landscape. It has become one of the most serious environmental problems that raised much interest throughout the world. Monitoring soil erosion risk is an important part of soil conservation practices. The estimation of soil loss risk is the first step for a successful control of water erosion. The aim of this study is to estimate the soil loss risk and its spatial distribution in the different fields of Morocco and to prioritize areas for soil conservation interventions. The approach followed is the Revised Universal Soil Loss Equation (RUSLE) using remote sensing and GIS, which is the most popular empirically based model used globally for erosion prediction and control. This model has been tested in many agricultural watersheds in the world, particularly for large-scale basins due to the simplicity of the model formulation and easy availability of the dataset. The spatial distribution of the annual soil loss was elaborated by the combination of several factors: rainfall erosivity, soil erodability, topography, and land cover. The average annual soil loss estimated in several basins watershed of Morocco varies from 0 to 50t/ha/year. Watersheds characterized by high-erosion-vulnerability are located in the North (Rif Mountains) and more particularly in the Central part of Morocco (High Atlas Mountains). This variation of vulnerability is highly correlated to slope variation which indicates that the topography factor is the main agent of soil erosion within these basin catchments. These results could be helpful for the planning of natural resources management and for implementing sustainable long-term management strategies which are necessary for soil conservation and for increasing over the projected economic life of the dam implemented.

Keywords: soil loss, RUSLE, GIS-remote sensing, watershed, Morocco

Procedia PDF Downloads 464

Authors: Sarah E. Nelson, Casey Weiner, Alexander Sigmon, Jun Hua, Haris I. Sair, Jose I. Suarez, Robert D. Stevens

Abstract:

Graph-theoretical information from structural connectomes indicated significant connectivity changes and improved acute prognostication in a Random Forest (RF) model in aneurysmal subarachnoid hemorrhage (aSAH), which can lead to significant morbidity and mortality and has traditionally been fraught by poor methods to predict outcome. This study’s hypothesis was that structural connectivity changes occur in canonical brain networks of acute aSAH patients, and that these changes are associated with functional outcome at six months. In a prospective cohort of patients admitted to a single institution for management of acute aSAH, patients underwent diffusion tensor imaging (DTI) as part of a multimodal MRI scan. A weighted undirected structural connectome was created of each patient’s images using Constant Solid Angle (CSA) tractography, with 176 regions of interest (ROIs) defined by the Johns Hopkins Eve atlas. ROIs were sorted into four networks: Default Mode Network, Executive Control Network, Salience Network, and Whole Brain. The resulting nodes and edges were characterized using graph-theoretic features, including Node Strength (NS), Betweenness Centrality (BC), Network Degree (ND), and Connectedness (C). Clinical (including demographics and World Federation of Neurologic Surgeons scale) and graph features were used separately and in combination to train RF and Logistic Regression classifiers to predict two outcomes: dichotomized modified Rankin Score (mRS) at discharge and at six months after discharge (favorable outcome mRS 0-2, unfavorable outcome mRS 3-6). A total of 56 aSAH patients underwent DTI a median (IQR) of 7 (IQR=8.5) days after admission. The best performing model (RF) combining clinical and DTI graph features had a mean Area Under the Receiver Operator Characteristic Curve (AUROC) of 0.88 ± 0.00 and Area Under the Precision Recall Curve (AUPRC) of 0.95 ± 0.00 over 500 trials. The combined model performed better than the clinical model alone (AUROC 0.81 ± 0.01, AUPRC 0.91 ± 0.00). The highest-ranked graph features for prediction were NS, BC, and ND. These results indicate reorganization of the connectome early after aSAH. The performance of clinical prognostic models was increased significantly by the inclusion of DTI-derived graph connectivity metrics. This methodology could significantly improve prognostication of aSAH.

Keywords: connectomics, diffusion tensor imaging, graph theory, machine learning, subarachnoid hemorrhage

Procedia PDF Downloads 190

Authors: Shengnan Chen, Shuhua Wang

Abstract:

Successful production of hydrocarbon from unconventional tight oil reserves has changed the energy landscape in North America. The oil contained within these reservoirs typically will not flow to the wellbore at economic rates without assistance from advanced horizontal well and multi-stage hydraulic fracturing. Efficient and economic development of these reserves is a priority of society, government, and industry, especially under the current low oil prices. Meanwhile, society needs technological and process innovations to enhance oil recovery while concurrently reducing environmental impacts. Recently, big data analysis and artificial intelligence become very popular, developing data-driven insights for better designs and decisions in various engineering disciplines. However, the application of data mining in petroleum engineering is still in its infancy. The objective of this research aims to apply intelligent data analysis and data-driven models to exploit unconventional oil reserves both efficiently and economically. More specifically, a comprehensive database including the reservoir geological data, reservoir geophysical data, well completion data and production data for thousands of wells is firstly established to discover the valuable insights and knowledge related to tight oil reserves development. Several data analysis methods are introduced to analysis such a huge dataset. For example, K-means clustering is used to partition all observations into clusters; principle component analysis is applied to emphasize the variation and bring out strong patterns in the dataset, making the big data easy to explore and visualize; exploratory factor analysis (EFA) is used to identify the complex interrelationships between well completion data and well production data. Different data mining techniques, such as artificial neural network, fuzzy logic, and machine learning technique are then summarized, and appropriate ones are selected to analyze the database based on the prediction accuracy, model robustness, and reproducibility. Advanced knowledge and patterned are finally recognized and integrated into a modified self-adaptive differential evolution optimization workflow to enhance the oil recovery and maximize the net present value (NPV) of the unconventional oil resources. This research will advance the knowledge in the development of unconventional oil reserves and bridge the gap between the big data and performance optimizations in these formations. The newly developed data-driven optimization workflow is a powerful approach to guide field operation, which leads to better designs, higher oil recovery and economic return of future wells in the unconventional oil reserves.

Keywords: big data, artificial intelligence, enhance oil recovery, unconventional oil reserves

Procedia PDF Downloads 285

Authors: Corinna Zeli, Silvia Calati, Marco Simeoni, Chiara Comastri

Abstract:

Developmental stuttering affects about 10% of preschool children; although the high percentage of natural recovery, a quarter of them will become an adult who stutters. An effective early intervention should help those children with high persistence risk for the future. The Psicodizione method for early stuttering is an Italian behavior indirect treatment for preschool children who stutter in which method parents act as good guides for communication, modeling their own fluency. In this study, we give a preliminary measure to evaluate the long-term effectiveness of Psicodizione method on stuttering preschool children with a high persistence risk. Among all Italian children treated with the Psicodizione method between 2018 and 2019, we selected 8 kids with at least 3 high risk persistence factors from the Illinois Prediction Criteria proposed by Yairi and Seery. The factors chosen for the selection were: one parent who stutters (1pt mother; 1.5pt father), male gender, ≥ 4 years old at onset; ≥ 12 months from onset of symptoms before treatment. For this study, the families were contacted after an average period of time of 14,7 months (range 3 - 26 months). Parental reports were gathered with a standard online questionnaire in order to obtain data reflecting fluency from a wide range of the children’s life situations. The minimum worthwhile outcome was set at "mild evidence" in a 5 point Likert scale (1 mild evidence- 5 high severity evidence). A second group of 6 children, among those treated with the Piscodizione method, was selected as high potential for spontaneous remission (low persistence risk). The children in this group had to fulfill all the following criteria: female gender, symptoms for less than 12 months (before treatment), age of onset <4 years old, none of the parents with persistent stuttering. At the time of this follow-up, the children were aged 6–9 years, with a mean of 15 months post-treatment. Among the children in the high persistence risk group, 2 (25%) hadn’t had stutter anymore, and 3 (37,5%) had mild stutter based on parental reports. In the low persistency risk group, the children were aged 4–6 years, with a mean of 14 months post-treatment, and 5 (84%) hadn’t had stutter anymore (for the past 16 months on average).62,5% of children at high risk of persistence after Psicodizione treatment showed mild evidence of stutter at most. 75% of parents confirmed a better fluency than before the treatment. The low persistence risk group seemed to be representative of spontaneous recovery. This study’s design could help to better evaluate the success of the proposed interventions for stuttering preschool children and provides a preliminary measure of the effectiveness of the Psicodizione method on high persistence risk children.

Keywords: early treatment, fluency, preschool children, stuttering

Procedia PDF Downloads 219

Authors: Fangliang Chen, Yihe Huang, Tejav Deganyar, Anselm Boehm, Hamid Batoul

Abstract:

With growing population density and further urbanization, the average noise level in cities is increasing. Excessive noise is not only annoying but also leads to a negative impact on human health. To deal with the increasing city noise, environmental regulations bring up higher standards on acoustic comfort in buildings by mitigating the noise transmission from building envelope exterior to interior. Framed window, door and façade systems are the leading choice for modern fenestration construction, which provides demonstrated quality of weathering reliability, environmental efficiency, and installation ease. The overall sound insulation of such systems depends both on glasses and frames, where glass usually covers the majority of the exposed surfaces, thus it is the main source of sound energy transmission. While frames in modern façade systems become slimmer for aesthetic appearance, which contribute to a minimal percentage of exposed surfaces. Nevertheless, frames might provide substantial transmission paths for sound travels through because of much less mass crossing the path, thus becoming more critical in limiting the acoustic performance of the whole system. There are various methodologies and numerical programs that can accurately predict the acoustic performance of either glasses or frames. However, due to the vast variance of size and dimension between frame and glass in the same system, there is no satisfactory theoretical approach or affordable simulation tool in current practice to access the over acoustic performance of a whole façade system. For this reason, laboratory test turns out to be the only reliable source. However, laboratory test is very time consuming and high costly, moreover different lab might provide slightly different test results because of varieties of test chambers, sample mounting, and test operations, which significantly constrains the early phase design of framed façade systems. To address this dilemma, this study provides an effective analytical methodology to predict the acoustic performance of framed façade systems, based on vast amount of acoustic test results on glass, frame and the whole façade system consist of both. Further test results validate the current model is able to accurately predict the overall sound transmission loss of a framed system as long as the acoustic behavior of the frame is available. Though the presented methodology is mainly developed from façade systems with aluminum frames, it can be easily extended to systems with frames of other materials such as steel, PVC or wood.

Keywords: city noise, building facades, sound mitigation, sound transmission loss, framed façade system

Procedia PDF Downloads 63