Search results for: single-phase models

Authors: Ammarah Irum, Muhammad Ali Tahir

Abstract:

Document-level sentiment analysis in Urdu is a challenging Natural Language Processing (NLP) task due to the difficulty of working with lengthy texts in a language with constrained resources. Deep learning models, which are complex neural network architectures, are well-suited to text-based applications in addition to data formats like audio, image, and video. To investigate the potential of deep learning for Urdu sentiment analysis, we implemented five different deep learning models, including Bidirectional Long Short Term Memory (BiLSTM), Convolutional Neural Network (CNN), Convolutional Neural Network with Bidirectional Long Short Term Memory (CNN-BiLSTM), and Bidirectional Encoder Representation from Transformer (BERT). In this study, we developed a hybrid deep learning model called BiLSTM-Single Layer Multi Filter Convolutional Neural Network (BiLSTM-SLMFCNN) by fusing BiLSTM and CNN architecture. The proposed and baseline techniques are applied on Urdu Customer Support data set and IMDB Urdu movie review data set by using pre-trained Urdu word embedding that are suitable for sentiment analysis at the document level. Results of these techniques are evaluated and our proposed model outperforms all other deep learning techniques for Urdu sentiment analysis. BiLSTM-SLMFCNN outperformed the baseline deep learning models and achieved 83%, 79%, 83% and 94% accuracy on small, medium and large sized IMDB Urdu movie review data set and Urdu Customer Support data set respectively.

Keywords: urdu sentiment analysis, deep learning, natural language processing, opinion mining, low-resource language

Procedia PDF Downloads 72

Authors: Hermann Kuehnle

Abstract:

Developments in ICT totally reshape manufacturing as machines, objects and equipment on the shop floors will be smart and online. Interactions with virtualizations and models of a manufacturing unit will appear exactly as interactions with the unit itself. These virtualizations may be driven by providers with novel ICT services on demand that might jeopardize even well established business models. Context aware equipment, autonomous orders, scalable machine capacity or networkable manufacturing unit will be the terminology to get familiar with in manufacturing and manufacturing management. Such newly appearing smart abilities with impact on network behavior, collaboration procedures and human resource development will make distributed manufacturing a preferred model to produce. Computing miniaturization and smart devices revolutionize manufacturing set ups, as virtualizations and atomization of resources unwrap novel manufacturing principles. Processes and resources obey novel specific laws and have strategic impact on manufacturing and major operational implications. Mechanisms from distributed manufacturing engaging interacting smart manufacturing units and decentralized planning and decision procedures already demonstrate important effects from this shift of focus towards collaboration and interoperability.

Keywords: autonomous unit, networkability, smart manufacturing unit, virtualization

Procedia PDF Downloads 526

Authors: Hayder H. Kareem, Shunqi Pan

Abstract:

The geological environment where the groundwater is collected represents the most important element that affects the behaviour of groundwater aquifer. As groundwater is a worldwide vital resource, it requires knowing the parameters that affect this source accurately so that the conceptualized mathematical models would be acceptable to the broadest ranges. Therefore, groundwater models have recently become an effective and efficient tool to investigate groundwater aquifer behaviours. Groundwater aquifer may contain aquitards, aquicludes, or interfaces within its geological formations. Aquitards and aquicludes have geological formations that forced the modellers to include those formations within the conceptualized groundwater models, while interfaces are commonly neglected from the conceptualization process because the modellers believe that the interface has no effect on aquifer behaviour. The current research highlights the impact of an interface existing in a real unconfined groundwater aquifer called Dibdibba, located in Al-Najaf City, Iraq where it has a river called the Euphrates River that passes through the eastern part of this city. Dibdibba groundwater aquifer consists of two types of soil layers separated by an interface soil layer. A groundwater model is built for Al-Najaf City to explore the impact of this interface. Calibration process is done using PEST 'Parameter ESTimation' approach and the best Dibdibba groundwater model is obtained. When the soil interface is conceptualized, results show that the groundwater tables are significantly affected by that interface through appearing dry areas of 56.24 km² and 6.16 km² in the upper and lower layers of the aquifer, respectively. The Euphrates River will also leak water into the groundwater aquifer of 7359 m³/day. While these results are changed when the soil interface is neglected where the dry area became 0.16 km², the Euphrates River leakage became 6334 m³/day. In addition, the conceptualized models (with and without interface) reveal different responses for the change in the recharge rates applied on the aquifer through the uncertainty analysis test. The aquifer of Dibdibba in Al-Najaf City shows a slight deficit in the amount of water supplied by the current pumping scheme and also notices that the Euphrates River suffers from stresses applied to the aquifer. Ultimately, this study shows a crucial need to represent the interface soil layer in model conceptualization to be the intended and future predicted behaviours more reliable for consideration purposes.

Keywords: Al-Najaf City, groundwater aquifer behaviour, groundwater modelling, interface soil layer, Visual MODFLOW

Procedia PDF Downloads 183

Authors: Irene Yi

Abstract:

Machine learning, natural language processing, and neural network models of language are becoming more and more prevalent in the fields of technology and linguistics today. Training data for machines are at best, large corpora of human literature and at worst, a reflection of the ugliness in society. Machines have been trained on millions of human books, only to find that in the course of human history, derogatory and sexist adjectives are used significantly more frequently when describing females in history and literature than when describing males. This is extremely problematic, both as training data, and as the outcome of natural language processing. As machines start to handle more responsibilities, it is crucial to ensure that they do not take with them historical sexist and misogynistic notions. This paper gathers data and algorithms from neural network models of language having to deal with syntax, semantics, sociolinguistics, and text classification. Results are significant in showing the existing intentional and unintentional misogynistic notions used to train machines, as well as in developing better technologies that take into account the semantics and syntax of text to be more mindful and reflect gender equality. Further, this paper deals with the idea of non-binary gender pronouns and how machines can process these pronouns correctly, given its semantic and syntactic context. This paper also delves into the implications of gendered grammar and its effect, cross-linguistically, on natural language processing. Languages such as French or Spanish not only have rigid gendered grammar rules, but also historically patriarchal societies. The progression of society comes hand in hand with not only its language, but how machines process those natural languages. These ideas are all extremely vital to the development of natural language models in technology, and they must be taken into account immediately.

Keywords: gendered grammar, misogynistic language, natural language processing, neural networks

Procedia PDF Downloads 120

Authors: Lilla Dorina Habsz, Marta Rado

Abstract:

Popularity is affected by a variety of factors in the primary school such as academic achievement and ethnicity. The main goal of our study was to analyse whether acting white exists in Hungarian elementary schools. In other words, we observed whether Roma students penalize those in-group members who obtain the high academic achievement. Furthermore, to show how popularity is influenced by changes in academic achievement in inter-ethnic relations. The empirical basis of our research was the 'competition and negative networks' longitudinal dataset, which was collected by the MTA TK 'Lendület' RECENS research group. This research followed 11 and 12-year old students for a two-year period. The survey was analysed using fixed and random effect models. Overall, we found a positive correlation between grades and popularity, but no evidence for the acting white effect. However, better grades were more positively evaluated within the majority group than within the minority group, which may further increase inequalities.

Keywords: academic achievement, elementary school, ethnicity, popularity

Procedia PDF Downloads 200

Authors: Alireza Fallahfard, Ludwig Vinches, Stephane Halle

Abstract:

In the workplace, the exposure level of airborne contaminants should be evaluated due to health and safety issues. It can be done by numerical models or experimental measurements, but the numerical approach can be useful when it is challenging to perform experiments. One of the simplest models is the well-mixed room (WMR) model, which has shown its usefulness to predict inhalation exposure in many situations. However, since the WMR is limited to gases and vapors, it cannot be used to predict exposure to aerosols. The main objective is to modify the WMR model to expand its application to exposure scenarios involving aerosols. To reach this objective, the standard WMR model has been modified to consider the deposition of particles by gravitational settling and Brownian and turbulent deposition. Three deposition models were implemented in the model. The time-dependent concentrations of airborne particles predicted by the model were compared to experimental results conducted in a 0.512 m3 chamber. Polystyrene particles of 1, 2, and 3 µm in aerodynamic diameter were generated with a nebulizer under two air changes per hour (ACH). The well-mixed condition and chamber ACH were determined by the tracer gas decay method. The mean friction velocity on the chamber surfaces as one of the input variables for the deposition models was determined by computational fluid dynamics (CFD) simulation. For the experimental procedure, the particles were generated until reaching the steady-state condition (emission period). Then generation stopped, and concentration measurements continued until reaching the background concentration (decay period). The results of the tracer gas decay tests revealed that the ACHs of the chamber were: 1.4 and 3.0, and the well-mixed condition was achieved. The CFD results showed the average mean friction velocity and their standard deviations for the lowest and highest ACH were (8.87 ± 0.36) ×10-2 m/s and (8.88 ± 0.38) ×10-2 m/s, respectively. The numerical results indicated the difference between the predicted deposition rates by the three deposition models was less than 2%. The experimental and numerical aerosol concentrations were compared in the emission period and decay period. In both periods, the prediction accuracy of the modified model improved in comparison with the classic WMR model. However, there is still a difference between the actual value and the predicted value. In the emission period, the modified WMR results closely follow the experimental data. However, the model significantly overestimates the experimental results during the decay period. This finding is mainly due to an underestimation of the deposition rate in the model and uncertainty related to measurement devices and particle size distribution. Comparing the experimental and numerical deposition rates revealed that the actual particle deposition rate is significant, but the deposition mechanisms considered in the model were ten times lower than the experimental value. Thus, particle deposition was significant and will affect the airborne concentration in occupational settings, and it should be considered in the airborne exposure prediction model. The role of other removal mechanisms should be investigated.

Keywords: aerosol, CFD, exposure assessment, occupational settings, well-mixed room model, zonal model

Procedia PDF Downloads 103

Authors: Yuhanis Yusof, Zuriani Mustaffa, Siti Sakira Kamaruddin

Abstract:

Data mining is a fundamental technique in identifying patterns from large data sets. The extracted facts and patterns contribute in various domains such as marketing, forecasting, and medical. Prior to that, data are consolidated so that the resulting mining process may be more efficient. This study investigates the effect of different data normalization techniques, which are Min-max, Z-score, and decimal scaling, on Swarm-based forecasting models. Recent swarm intelligence algorithms employed includes the Grey Wolf Optimizer (GWO) and Artificial Bee Colony (ABC). Forecasting models are later developed to predict the daily spot price of crude oil and gasoline. Results showed that GWO works better with Z-score normalization technique while ABC produces better accuracy with the Min-Max. Nevertheless, the GWO is more superior that ABC as its model generates the highest accuracy for both crude oil and gasoline price. Such a result indicates that GWO is a promising competitor in the family of swarm intelligence algorithms.

Keywords: artificial bee colony, data normalization, forecasting, Grey Wolf optimizer

Procedia PDF Downloads 475

Authors: Jakub Zembrzuski, Bartosz Sobczyk, Mikołaj MIśkiewicz

Abstract:

Besides designing new constructions, engineers all over the world must face another problem – maintenance, repairs, and assessment of the technical condition of existing bridges. To solve more complex issues, it is necessary to be familiar with the theory of finite element method and to have access to the software that provides sufficient tools which to enable create of sometimes significantly advanced numerical models. The paper includes a brief assessment of the technical condition, a description of the in situ non-destructive testing carried out and the FEM models created for global and local analysis. In situ testing was performed using strain gauges and displacement sensors. Numerical models were created using various software and numerical modeling techniques. Particularly noteworthy is the method of modeling riveted joints of the crossbeam of the viaduct. It is a simplified method that consists of the use of only basic numerical tools such as beam and shell finite elements, constraints, and simplified boundary conditions (fixed support and symmetry). The results of the numerical analyses were presented and discussed. It is clearly explained why the structure did not fail, despite the fact that the weld of the deck plate completely failed. A further research problem that was solved was to determine the cause of the rapid increase in values on the stress diagram in the cross-section of the transverse section. The problems were solved using the solely mentioned, simplified method of modeling riveted joints, which demonstrates that it is possible to solve such problems without access to sophisticated software that enables to performance of the advanced nonlinear analysis. Moreover, the obtained results are of great importance in the field of assessing the operation of bridge structures with an orthotropic plate.

Keywords: bridge, diagnostics, FEM simulations, failure, NDT, in situ testing

Procedia PDF Downloads 72

Authors: Elizaveta S. Prokofyeva, Svetlana V. Maltseva, Roman D. Zaitsev

Abstract:

Rapidly evolving modern data analysis technologies in healthcare play a large role in understanding the operation of the system and its characteristics. Nowadays, one of the key tasks in urban healthcare is to optimize the resource allocation. Thus, the application of data analysis in medical institutions to solve optimization problems determines the significance of this study. The purpose of this research was to establish the dependence between the indicators of the effectiveness of the medical institution and its resources. Hospital discharges by diagnosis; hospital days of in-patients and in-patient average length of stay were selected as the performance indicators and the demand of the medical facility. The hospital beds by type of care, medical technology (magnetic resonance tomography, gamma cameras, angiographic complexes and lithotripters) and physicians characterized the resource provision of medical institutions for the developed models. The data source for the research was an open database of the statistical service Eurostat. The choice of the source is due to the fact that the databases contain complete and open information necessary for research tasks in the field of public health. In addition, the statistical database has a user-friendly interface that allows you to quickly build analytical reports. The study provides information on 28 European for the period from 2007 to 2016. For all countries included in the study, with the most accurate and complete data for the period under review, predictive models were developed based on historical panel data. An attempt to improve the quality and the interpretation of the models was made by cluster analysis of the investigated set of countries. The main idea was to assess the similarity of the joint behavior of the variables throughout the time period under consideration to identify groups of similar countries and to construct the separate regression models for them. Therefore, the original time series were used as the objects of clustering. The hierarchical agglomerate algorithm k-medoids was used. The sampled objects were used as the centers of the clusters obtained, since determining the centroid when working with time series involves additional difficulties. The number of clusters used the silhouette coefficient. After the cluster analysis it was possible to significantly improve the predictive power of the models: for example, in the one of the clusters, MAPE error was only 0,82%, which makes it possible to conclude that this forecast is highly reliable in the short term. The obtained predicted values of the developed models have a relatively low level of error and can be used to make decisions on the resource provision of the hospital by medical personnel. The research displays the strong dependencies between the demand for the medical services and the modern medical equipment variable, which highlights the importance of the technological component for the successful development of the medical facility. Currently, data analysis has a huge potential, which allows to significantly improving health services. Medical institutions that are the first to introduce these technologies will certainly have a competitive advantage.

Keywords: data analysis, demand modeling, healthcare, medical facilities

Procedia PDF Downloads 144

Authors: M. S. Wickramarachchi, L. S. Nawarathna, C. M. B. Dematawewa

Abstract:

Hatchery performance is critical for the profitability of poultry breeder operations. Some extrinsic parameters of eggs and breeders cause to increase or decrease the hatchability. This study aims to identify the affecting extrinsic parameters on the commercial hatchability of local chicken's eggs and determine the most efficient classification model with a hatchability rate greater than 90%. In this study, seven extrinsic parameters were considered: egg weight, moisture loss, breeders age, number of fertilised eggs, shell width, shell length, and shell thickness. Multiple linear regression was performed to determine the most influencing variable on hatchability. First, the correlation between each parameter and hatchability were checked. Then a multiple regression model was developed, and the accuracy of the fitted model was evaluated. Linear Discriminant Analysis (LDA), Classification and Regression Trees (CART), k-Nearest Neighbors (kNN), Support Vector Machines (SVM) with a linear kernel, and Random Forest (RF) algorithms were applied to classify the hatchability. This grouping process was conducted using binary classification techniques. Hatchability was negatively correlated with egg weight, breeders' age, shell width, shell length, and positive correlations were identified with moisture loss, number of fertilised eggs, and shell thickness. Multiple linear regression models were more accurate than single linear models regarding the highest coefficient of determination (R²) with 94% and minimum AIC and BIC values. According to the classification results, RF, CART, and kNN had performed the highest accuracy values 0.99, 0.975, and 0.972, respectively, for the commercial hatchery process. Therefore, the RF is the most appropriate machine learning algorithm for classifying the breeder outcomes, which are economically profitable or not, in a commercial hatchery.

Keywords: classification models, egg weight, fertilised eggs, multiple linear regression

Procedia PDF Downloads 87

Authors: Saif Obaid Alkaabi, Nabil Ayad

Abstract:

Governments constantly seek to offer faster, more secure, efficient and effective services for their citizens. Recent changes and developments to communication services and technologies, mainly due the Internet, have led to immense improvements in the way governments of advanced countries carry out their interior operations Therefore, advances in e-government services have been broadly adopted and used in various developed countries, as well as being adapted to developing countries. The implementation of advances depends on the utilization of the most innovative structures of data techniques, mainly in web dependent applications, to enhance the main functions of governments. These functions, in turn, have spread to mobile and wireless techniques, generating a new advanced direction called m-government. This paper discusses a selection of available m-government applications and several business modules and frameworks in various fields. Practically, the m-government models, techniques and methods have become the improved version of e-government. M-government offers the potential for applications which will work better, providing citizens with services utilizing mobile communication and data models incorporating several government entities. Developing countries can benefit greatly from this innovation due to the fact that a large percentage of their population is young and can adapt to new technology and to the fact that mobile computing devices are more affordable. The use of models of mobile transactions encourages effective participation through the use of mobile portals by businesses, various organizations, and individual citizens. Although the application of m-government has great potential, it does have major limitations. The limitations include: the implementation of wireless networks and relative communications, the encouragement of mobile diffusion, the administration of complicated tasks concerning the protection of security (including the ability to offer privacy for information), and the management of the legal issues concerning mobile applications and the utilization of services.

Keywords: e-government, m-government, system dependability, system security, trust

Procedia PDF Downloads 381

Authors: Ivan Sosa Gomez

Abstract:

Digitalization has impacted the entire insurance value chain, and the growing movement towards P2P platforms and the collaborative economy is also beginning to have a significant impact. P2P insurance is defined as innovation, enabling policyholders to pool their capital, self-organize, and self-manage their own insurance. In this context, new InsurTech start-ups are emerging as peer-to-peer (P2P) providers, based on a model that differs from traditional insurance. As a result, although P2P platforms do not change the fundamental basis of insurance, they do enable potentially more efficient business models to be established in terms of ensuring the coverage of risk. It is therefore relevant to determine whether p2p innovation can have substantial effects on the future of the insurance sector. For this purpose, it is considered necessary to develop P2P innovation from a business perspective, as well as to build a comparison between a traditional model and a P2P model from an actuarial perspective. Objectives: The objectives are (1) to represent P2P innovation in the business model compared to the traditional insurance model and (2) to establish a comparison between a traditional model and a P2P model from an actuarial perspective. Methodology: The research design is defined as action research in terms of understanding and solving the problems of a collectivity linked to an environment, applying theory and best practices according to the approach. For this purpose, the study is carried out through the participatory variant, which involves the collaboration of the participants, given that in this design, participants are considered experts. For this purpose, prolonged immersion in the field is carried out as the main instrument for data collection. Finally, an actuarial model is developed relating to the calculation of premiums that allows for the establishment of projections of future scenarios and the generation of conclusions between the two models. Main Contributions: From an actuarial and business perspective, we aim to contribute by developing a comparison of the two models in the coverage of risk in order to determine whether P2P innovation can have substantial effects on the future of the insurance sector.

Keywords: Insurtech, innovation, business model, P2P, insurance

Procedia PDF Downloads 92

Authors: Behzad Behnia, Noah LaRussa-Trott

Abstract:

In recent years, fibers have been successfully used as an additive to reinforce asphalt concrete materials and to enhance the sustainability and resiliency of transportation infrastructure. Roads covered with fiber-reinforced asphalt concrete (FRAC) require less frequent maintenance and tend to have a longer lifespan. The present work investigates the application of sasobit-coated aramid fibers in asphalt pavements and employs machine learning to develop prediction models to evaluate the cracking performance of FRAC materials. For the experimental part of the study, the effects of several important parameters such as fiber content, fiber length, and testing temperature on fracture characteristics of FRAC mixtures were thoroughly investigated. Two mechanical performance tests, i.e., the disk-shaped compact tension [DC(T)] and indirect tensile [ID(T)] strength tests, as well as the non-destructive acoustic emission test, were utilized to experimentally measure the cracking behavior of the FRAC material in both macro and micro level, respectively. The experimental results were used to train the supervised machine learning approach in order to establish prediction models for fracture performance of the FRAC mixtures in the field. Experimental results demonstrated that adding fibers improved the overall fracture performance of asphalt concrete materials by increasing their fracture energy, tensile strength and lowering their 'embrittlement temperature'. FRAC mixtures containing long-size fibers exhibited better cracking performance than regular-size fiber mixtures. The developed prediction models of this study could be easily employed by pavement engineers in the assessment of the FRAC pavements.

Keywords: fiber reinforced asphalt concrete, machine learning, cracking performance tests, prediction model

Procedia PDF Downloads 141

Authors: Parminder Kaur Heer, Alexei Lapkin

Abstract:

Due to global challenges of increased competition and demand for more sustainable products/processes, there is a rising pressure on the industry to develop innovative processes. Through Process Intensification (PI) the existing and new processes may be able to attain higher efficiency. However, very few PI options are generally considered. This is because processes are typically analysed at a unit operation level, thus limiting the search space for potential process options. PI performed at more detailed levels of a process can increase the size of the search space. The different levels at which PI can be achieved is unit operations, functional and phenomena level. Physical/chemical phenomena form the lowest level of aggregation and thus, are expected to give the highest impact because all the intensification options can be described by their enhancement. The objective of the current work is thus, generation of numerous process alternatives based on phenomena, and development of their corresponding computer aided models. The methodology comprises: a) automated generation of process options, and b) automated generation of process models. The process under investigation is disintegrated into functions viz. reaction, separation etc., and these functions are further broken down into the phenomena required to perform them. E.g., separation may be performed via vapour-liquid or liquid-liquid equilibrium. A list of phenomena for the process is formed and new phenomena, which can overcome the difficulties/drawbacks of the current process or can enhance the effectiveness of the process, are added to the list. For instance, catalyst separation issue can be handled by using solid catalysts; the corresponding phenomena are identified and added. The phenomena are then combined to generate all possible combinations. However, not all combinations make sense and, hence, screening is carried out to discard the combinations that are meaningless. For example, phase change phenomena need the co-presence of the energy transfer phenomena. Feasible combinations of phenomena are then assigned to the functions they execute. A combination may accomplish a single or multiple functions, i.e. it might perform reaction or reaction with separation. The combinations are then allotted to the functions needed for the process. This creates a series of options for carrying out each function. Combination of these options for different functions in the process leads to the generation of superstructure of process options. These process options, which are formed by a list of phenomena for each function, are passed to the model generation algorithm in the form of binaries (1, 0). The algorithm gathers the active phenomena and couples them to generate the model. A series of models is generated for the functions, which are combined to get the process model. The most promising process options are then chosen subjected to a performance criterion, for example purity of product, or via a multi-objective Pareto optimisation. The methodology was applied to a two-step process and the best route was determined based on the higher product yield. The current methodology can identify, produce and evaluate process intensification options from which the optimal process can be determined. It can be applied to any chemical/biochemical process because of its generic nature.

Keywords: Phenomena, Process intensification, Process models , Process options

Procedia PDF Downloads 232

Authors: Zhaoxin Luo, Michael Zhu

Abstract:

In natural languages, there are always complex semantic hierarchies. Obtaining the feature representation based on these complex semantic hierarchies becomes the key to the success of the model. Several RNN models have recently been proposed to use latent indicators to obtain the hierarchical structure of documents. However, the model that only uses a single-layer latent indicator cannot achieve the true hierarchical structure of the language, especially a complex language like Chinese. In this paper, we propose a deep layered model that stacks arbitrarily many RNN layers equipped with latent indicators. After using EM and training it hierarchically, our model solves the computational problem of stacking RNN layers and makes it possible to stack arbitrarily many RNN layers. Our deep hierarchical model not only achieves comparable results to large pre-trained models on the Chinese short text classification problem but also achieves state of art results on the Chinese long text classification problem.

Keywords: nature language processing, recurrent neural network, hierarchical structure, document classification, Chinese

Procedia PDF Downloads 68

Authors: Layan Moussa, Darine Salam, Samir Mustapha

Abstract:

Heavy metal contamination in agricultural soils in Lebanon poses serious environmental and health problems. Intensive efforts are employed to improve existing quantification methods of heavy metals in contaminated environments since conventional detection techniques have shown to be time-consuming, tedious, and costly. The implication of hyperspectral remote sensing in this field is possible and promising. However, factors impacting the efficiency of hyperspectral imaging in detecting and quantifying heavy metals in agricultural soils were not thoroughly studied. This study proposes to assess the use of hyperspectral imaging for the detection of Ni in agricultural clay soil collected from the Bekaa Valley, a major agricultural area in Lebanon, under different contamination levels and soil moisture content. Soil samples were contaminated with Ni, with concentrations ranging from 150 mg/kg to 4000 mg/kg. On the other hand, soil with background contamination was subjected to increased moisture levels varying from 5 to 75%. Hyperspectral imaging was used to detect and quantify Ni contamination in the soil at different contamination levels and moisture content. IBM SPSS statistical software was used to develop models that predict the concentration of Ni and moisture content in agricultural soil. The models were constructed using linear regression algorithms. The spectral curves obtained reflected an inverse correlation between both Ni concentration and moisture content with respect to reflectance. On the other hand, the models developed resulted in high values of predicted R2 of 0.763 for Ni concentration and 0.854 for moisture content. Those predictions stated that Ni presence was well expressed near 2200 nm and that of moisture was at 1900 nm. The results from this study would allow us to define the potential of using the hyperspectral imaging (HSI) technique as a reliable and cost-effective alternative for heavy metal pollution detection in contaminated soils and soil moisture prediction.

Keywords: heavy metals, hyperspectral imaging, moisture content, soil contamination

Procedia PDF Downloads 101

Authors: Yinheng Li

Abstract:

The remarkable advancements in large language models (LLMs) have brought about significant improvements in natural language processing tasks. This paper presents a comprehensive review of in-context learning techniques, focusing on different types of prompts, including discrete, continuous, few-shot, and zero-shot, and their impact on LLM performance. We explore various approaches to prompt design, such as manual design, optimization algorithms, and evaluation methods, to optimize LLM performance across diverse tasks. Our review covers key research studies in prompt engineering, discussing their methodologies and contributions to the field. We also delve into the challenges faced in evaluating prompt performance, given the absence of a single ”best” prompt and the importance of considering multiple metrics. In conclusion, the paper highlights the critical role of prompt design in harnessing the full potential of LLMs and provides insights into the combination of manual design, optimization techniques, and rigorous evaluation for more effective and efficient use of LLMs in various Natural Language Processing (NLP) tasks.

Keywords: in-context learning, prompt engineering, zero-shot learning, large language models

Procedia PDF Downloads 81

Authors: Mark Schudeleit, Christian Sieg, Ferit Küçükay

Abstract:

This paper describes how to dimension the electric components of a 48V hybrid system considering real customer use. Furthermore, it provides information about savings in energy and CO2 emissions by a customer-tailored 48V hybrid. Based on measured customer profiles, the electric units such as the electric motor and the energy storage are dimensioned. Furthermore, the CO2 reduction potential in real customer use is determined compared to conventional vehicles. Finally, investigations are carried out to specify the topology design and preliminary considerations in order to hybridize a conventional vehicle with a 48V hybrid system. The emission model results from an empiric approach also taking into account the effects of engine dynamics on emissions. We analyzed transient engine emissions during representative customer driving profiles and created emission meta models. The investigation showed a significant difference in emissions when simulating realistic customer driving profiles using the created verified meta models compared to static approaches which are commonly used for vehicle simulation.

Keywords: customer use, dimensioning, hybrid electric vehicles, vehicle simulation, 48V hybrid system

Procedia PDF Downloads 507

Authors: Shanxiong Chen, Xu Han, Xiaolong Wang, Hui Ma

Abstract:

Yi is an ethnic group mainly living in mainland China, with its own spoken and written language systems, after development of thousands of years. Ancient Yi is one of the six ancient languages in the world, which keeps a record of the history of the Yi people and offers documents valuable for research into human civilization. Recognition of the characters in ancient Yi helps to transform the documents into an electronic form, making their storage and spreading convenient. Due to historical and regional limitations, research on recognition of ancient characters is still inadequate. Thus, deep learning technology was applied to the recognition of such characters. Five models were developed on the basis of the four-layer convolutional neural network (CNN). Alpha-Beta divergence was taken as a penalty term to re-encode output neurons of the five models. Two fully connected layers fulfilled the compression of the features. Finally, at the softmax layer, the orthographic features of ancient Yi characters were re-evaluated, their probability distributions were obtained, and characters with features of the highest probability were recognized. Tests conducted show that the method has achieved higher precision compared with the traditional CNN model for handwriting recognition of the ancient Yi.

Keywords: recognition, CNN, Yi character, divergence

Procedia PDF Downloads 164

Authors: Ritu Raj, Hrishikesh Dubey

Abstract:

This paper presents the results of the experimental investigations carried out on two intricate plan shaped buildings to evaluate aerodynamic performance of the building. The purpose is to study the associated environment arising due to wind forces in isolated and interference conditions on a model of scale 1:300 with a prototype having 180m height. Experimental tests were carried out at the boundary layer wind tunnel considering isolated conditions with 0° to 180° isolated wind directions and four interference conditions of twin building (separately for both the models). The research has been undertaken in Terrain Category-II, which is the most widely available terrain in India. A comparative assessment of the two models is performed out in an attempt to comprehend the various consequences of diverse conditions that may emerge in real-life situations, as well as the discrepancies amongst them. Experimental results of wind pressure coefficients of Model-1 and Model-2 shows good agreement with various wind incidence conditions with minute difference in the magnitudes of mean Cp. On the basis of wind tunnel studies, it is distinguished that the performance of Model-2 is better than Model-1in both isolated as well as interference conditions for all wind incidences and orientations respectively.

Keywords: interference factor, tall buildings, wind direction, mean pressure-coefficients

Procedia PDF Downloads 128

Authors: Yujie Lei

Abstract:

This article examines the evolving concept of urban living through the lens of co-living spaces, focusing on Liverpool. It explores how co-living can address challenges such as rising urban isolation, housing affordability, and social autism, particularly among younger generations. The research aims to understand how these spaces can mitigate social isolation and maximize urban space use. Using a case study approach, the study examines models like Superloft, co-office spaces, and platforms like Airbnb. Findings reveal that Liverpool’s co-living initiatives have gained popularity, offering flexibility and community engagement. This concept has the potential for expansion, not only for the younger generation but also for elderly communities, fostering intergenerational living. The dissertation concludes that co-living offers a sustainable alternative to traditional housing models, aligning with digital-age lifestyles that prioritize flexibility and community. It presents a promising framework for shaping the future of urban development.

Keywords: co-living, urban design, social isolation, urban development, housing challenges

Procedia PDF Downloads 26

Authors: Redvan Ghasemlounia, Elham Ansari, Hikmet Kerem Cigizoglu

Abstract:

Hydrological modeling in arid and semi-arid regions is very important. Iran has many regions with these climate conditions such as Chaharmahal and Bakhtiari province that needs lots of attention with an appropriate management. Forecasting of hydrological parameters and estimation of hydrological events of catchments, provide important information that used for design, management and operation of water resources such as river systems, and dams, widely. Discharge in rivers is one of these parameters. This study presents the application and comparison of some estimation methods such as Feed-Forward Back Propagation Neural Network (FFBPNN), Multi Linear Regression (MLR), Gene Expression Programming (GEP) and Bayesian Network (BN) to predict the daily flow discharge of the Soolegan River, located at Chaharmahal and Bakhtiari province, in Iran. In this study, Soolegan, station was considered. This Station is located in Soolegan River at 51° 14՜ Latitude 31° 38՜ longitude at North Karoon basin. The Soolegan station is 2086 meters higher than sea level. The data used in this study are daily discharge and daily precipitation of Soolegan station. Feed Forward Back Propagation Neural Network(FFBPNN), Multi Linear Regression (MLR), Gene Expression Programming (GEP) and Bayesian Network (BN) models were developed using the same input parameters for Soolegan's daily discharge estimation. The results of estimation models were compared with observed discharge values to evaluate performance of the developed models. Results of all methods were compared and shown in tables and charts.

Keywords: ANN, multi linear regression, Bayesian network, forecasting, discharge, gene expression programming

Procedia PDF Downloads 561

Authors: Lara Bettinelli, Bernhard Glatz, Josef Fink

Abstract:

Planning engineers and researchers use various calculation models with different levels of complexity, calculation efficiency and accuracy in dynamic calculations of railway bridges under high-speed traffic. When choosing a vehicle model to depict the dynamic loading on the bridge structure caused by passing high-speed trains, different goals are pursued: On the one hand, the selected vehicle models should allow the calculation of a bridge’s vibrations as realistic as possible. On the other hand, the computational efficiency and manageability of the models should be preferably high to enable a wide range of applications. The commonly adopted and straightforward vehicle model is the moving load model (MLM), which simplifies the train to a sequence of static axle loads moving at a constant speed over the structure. However, the MLM can significantly overestimate the structure vibrations, especially when resonance events occur. More complex vehicle models, which depict the train as a system of oscillating and coupled masses, can reproduce the interaction dynamics between the vehicle and the bridge superstructure to some extent and enable the calculation of more realistic bridge accelerations. At the same time, such multi-body models require significantly greater processing capacities and precise knowledge of various vehicle properties. The European standards allow for applying the so-called additional damping method when simple load models, such as the MLM, are used in dynamic calculations. An additional damping factor depending on the bridge span, which should take into account the vibration-reducing benefits of the vehicle-bridge interaction, is assigned to the supporting structure in the calculations. However, numerous studies show that when the current standard specifications are applied, the calculation results for the bridge accelerations are in many cases still too high compared to the measured bridge accelerations, while in other cases, they are not on the safe side. A proposal to calculate the additional damping based on extensive dynamic calculations for a parametric field of simply supported bridges with a ballasted track was developed to address this issue. In this contribution, several different approaches to determine the additional damping of the supporting structure considering the vehicle-bridge interaction when using the MLM are compared with one another. Besides the standard specifications, this includes the approach mentioned above and two additional recently published alternative formulations derived from analytical approaches. For a bridge catalogue of 65 existing bridges in Austria in steel, concrete or composite construction, calculations are carried out with the MLM for two different high-speed trains and the different approaches for additional damping. The results are compared with the calculation results obtained by applying a more sophisticated multi-body model of the trains used. The evaluation and comparison of the results allow assessing the benefits of different calculation concepts for the additional damping regarding their accuracy and possible applications. The evaluation shows that by applying one of the recently published redesigned additional damping methods, the calculation results can reflect the influence of the vehicle-bridge interaction on the design-relevant structural accelerations considerably more reliable than by using normative specifications.

Keywords: Additional Damping Method, Bridge Dynamics, High-Speed Railway Traffic, Vehicle-Bridge-Interaction

Procedia PDF Downloads 161

Authors: Kaustubh Chakradeo, Michael Delves, Sofya Titarenko

Abstract:

Malaria is a serious disease which affects hundreds of millions of people around the world, each year. If not treated in time, it can be fatal. Despite recent developments in malaria diagnostics, the microscopy method to detect malaria remains the most common. Unfortunately, the accuracy of microscopic diagnostics is dependent on the skill of the microscopist and limits the throughput of malaria diagnosis. With the development of Artificial Intelligence tools and Deep Learning techniques in particular, it is possible to lower the cost, while achieving an overall higher accuracy. In this paper, we present a VGG-based model and compare it with previously developed models for identifying infected cells. Our model surpasses most previously developed models in a range of the accuracy metrics. The model has an advantage of being constructed from a relatively small number of layers. This reduces the computer resources and computational time. Moreover, we test our model on two types of datasets and argue that the currently developed deep-learning-based methods cannot efficiently distinguish between infected and contaminated cells. A more precise study of suspicious regions is required.

Keywords: convolution neural network, deep learning, malaria, thin blood smears

Procedia PDF Downloads 130

Authors: Sedat Yayla, Azhin Abdullah

Abstract:

In the process of constructing hydroelectric power plants, the Pelton turbine, which is characterized by its simple manufacturing and construction, is performed in high head and low water flow. Parameters of the turbine have to be comprised in the designing process for obtaining hydraulic turbine with the highest efficiency during different operating conditions. The present investigation applied three-dimensional computational fluid dynamics (CFD). In addition, the bucket of Pelton turbine models with different splitter angle and inlet velocity values were examined for determining the force and visualizing the flow pattern on the bucket. The study utilized two diverse bucket models at various inlet velocities (20, 25, 30,35and 40m/s) and four different splitter angles (55, 75,90and 115 degree) for finding out the impacts of every single parameter on the effective force on the bucket. The acquired outcomes revealed that there is a linear relationship between force and inlet velocity on the bucket. Furthermore, the results also uncovered that the relationship between splitter angle and force on the bucket is linear until 90 degree.

Keywords: bucket design, computational fluid dynamics (CFD), free surface flow, two-phase flow, volume of fluid (VOF)

Procedia PDF Downloads 271

Authors: Raghabendra Yadav, Baochun Chen, Huihui Yuan, Zhibin Lian

Abstract:

Concrete Filled Steel Tube (CFST) columns are widely used in Civil Engineering Structures due to their abundant properties. CFST-RC column is a built up column in which CFST members are connected with RC web. The CFST-RC column has excellent static and earthquake resistant properties, such as high strength, high ductility and large energy absorption capacity. CFST-RC columns have been adopted as piers in Ganhaizi Bridge in high seismic risk zone with a highest pier of 107m. The experimental investigation on scaled models of similar type of the CFST-RC pier are carried out. The experimental investigation on scaled models of similar type of the CFST-RC pier are carried out. Under cyclic loading, the hysteretic performance of CFST-RC columns, such as failure modes, ductility, load displacement hysteretic curves, energy absorption capacity, strength and stiffness degradation are studied in this paper.

Keywords: CFST, cyclic load, Ganhaizi bridge, seismic performance

Procedia PDF Downloads 246

Authors: Yaowei Hu, Yongkai Wu, Lu Zhang, Xintao Wu

Abstract:

Causal inference from observational data is receiving wide applications in many fields. However, unidentifiable situations, where causal effects cannot be uniquely computed from observational data, pose critical barriers to applying causal inference to complicated real applications. In this paper, we develop a bounding method for estimating the average causal effect (ACE) under unidentifiable situations due to hidden confounders. We propose to parameterize the unknown exogenous random variables and structural equations of a causal model using neural networks and implicit generative models. Then, with an adversarial learning framework, we search the parameter space to explicitly traverse causal models that agree with the given observational distribution and find those that minimize or maximize the ACE to obtain its lower and upper bounds. The proposed method does not make any assumption about the data generating process and the type of the variables. Experiments using both synthetic and real-world datasets show the effectiveness of the method.

Keywords: average causal effect, hidden confounding, bound estimation, generative adversarial learning

Procedia PDF Downloads 191

Authors: Redouane Tlemsani, Redouane, Belkacem Kouninef, Abdelkader Benyettou

Abstract:

In this paper, our system is a Markovien system which we can see it like a Dynamic Bayesian Networks. One of the major interests of these systems resides in the complete training of the models (topology and parameters) starting from training data. The Bayesian Networks are representing models of dubious knowledge on complex phenomena. They are a union between the theory of probability and the graph theory in order to give effective tools to represent a joined probability distribution on a set of random variables. The representation of knowledge bases on description, by graphs, relations of causality existing between the variables defining the field of study. The theory of Dynamic Bayesian Networks is a generalization of the Bayesians networks to the dynamic processes. Our objective amounts finding the better structure which represents the relationships (dependencies) between the variables of a dynamic bayesian network. In applications in pattern recognition, one will carry out the fixing of the structure which obliges us to admit some strong assumptions (for example independence between some variables).

Keywords: Arabic on line character recognition, dynamic Bayesian network, pattern recognition, networks

Procedia PDF Downloads 617

Authors: Akriti Chandra, Gourav Dwivedi, Seema Sharma, Shivani

Abstract:

This study investigates the adoption of green business (GB) models within a circular economy framework (CEBM) for Micro Small and Medium Enterprise (MSME), given the rising importance of sustainable practices. The research begins by exploring the shift from linear business models towards resource-efficient, sustainable models, emphasizing the benefits of the circular economy. The study's literature review identifies 60 influential factors impacting the shift to green businesses, grouped as internal and external drivers. However, there is a research gap in examining these factors' interrelationships and operationalizing them within MSMEs. To address this gap, the study employs Total Interpretive Structural Modelling (TISM) to establish a hierarchical structure of factors influencing GB and circular economy business model (CEBM) adoption. Findings reveal that factors like green innovation and market competitiveness are particularly impactful. Using Systems Theory, which views organizations as complex adaptive systems, the study contextualizes these drivers within MSMEs, proposing a framework for a sustainable business model adoption. The study concludes with significant implications for policymakers, suggesting that the identified factors and their hierarchical relationships can guide policy formulation for a broader transition to green business practices. This work also invites further research, recommending larger, quantitative studies to empirically validate these factors and explore practical challenges in implementing CEBMs.

Keywords: green business (GB), circular economy business model (CEBM), micro small and medium enterprise (MSME), total interpretive structural modelling (TISM), systems theory

Procedia PDF Downloads 12

Authors: Farid Gaci, Zoubir Nemouchi

Abstract:

A numerical study of laminar and turbulent fluid flows in 90° bend of square section was carried out. Three-dimensional meshes, based on hexahedral cells, were generated. The QUICK scheme was employed to discretize the convective term in the transport equations. The SIMPLE algorithm was adopted to treat the velocity-pressure coupling. The flow structure obtained showed interesting features such as recirculation zones and counter-rotating pairs of vortices. The performance of three different turbulence models was evaluated: the standard k- ω model, the SST k-ω model and the Reynolds Stress Model (RSM). Overall, it was found that, the multi-equation model performed better than the two equation models. In fact, the existence of four pairs of counter rotating cells, in the straight duct upstream of the bend, were predicted by the RSM closure but not by the standard eddy viscosity model nor the SST k-ω model. The analysis of the results led to a better understanding of the induced three dimensional secondary flows and the behavior of the local pressure coefficient and the friction coefficient.

Keywords: curved duct, counter-rotating cells, secondary flow, laminar, turbulent

Procedia PDF Downloads 336