Search results for: binary vector quantization (BVQ)

Authors: Ganfeng Yu, Yuefeng Ma, Shanliang Yang

Abstract:

The word discovery is a key problem in text information retrieval technology. Methods in new word discovery tend to be closely related to words because they generally obtain new word results by analyzing words. With the popularity of social networks, individual netizens and online self-media have generated various network texts for the convenience of online life, including network words that are far from standard Chinese expression. How detect network words is one of the important goals in the field of text information retrieval today. In this paper, we integrate the word embedding model and clustering methods to propose a network word discovery framework based on sentence semantic similarity (S³-NWD) to detect network words effectively from the corpus. This framework constructs sentence semantic vectors through a distributed representation model, uses the similarity of sentence semantic vectors to determine the semantic relationship between sentences, and finally realizes network word discovery by the meaning of semantic replacement between sentences. The experiment verifies that the framework not only completes the rapid discovery of network words but also realizes the standard word meaning of the discovery of network words, which reflects the effectiveness of our work.

Keywords: text information retrieval, natural language processing, new word discovery, information extraction

Procedia PDF Downloads 65

Authors: Md Rafiqul Islam, Ashir Ahmed, Anwaar Ulhaq, Abu Raihan M. Kamal, Yuan Miao, Hua Wang

Abstract:

Mental health of people is equally important as of their physical health. Mental health and well-being are influenced not only by individual attributes but also by the social circumstances in which people find themselves and the environment in which they live. Like physical health, there is a number of internal and external factors such as biological, social and occupational factors that could influence the mental health of people. People living in poverty, suffering from chronic health conditions, minority groups, and those who exposed to/or displaced by war or conflict are generally more likely to develop mental health conditions. However, to authors’ best knowledge, there is dearth of knowledge on the impact of workplace (especially the highly stressed IT/Tech workplace) on the mental health of its workers. This study attempts to examine the factors influencing the mental health of tech workers. A publicly available dataset containing more than 65,000 cells and 100 attributes is examined for this purpose. Number of machine learning techniques such as ‘Decision Tree’, ‘K nearest neighbor’ ‘Support Vector Machine’ and ‘Ensemble’, are then applied to the selected dataset to draw the findings. It is anticipated that the analysis reported in this study would contribute in presenting useful insights on the attributes contributing in the mental health of tech workers using relevant machine learning techniques.

Keywords: mental disorder, diagnosis, occupational stress, IT workplace

Procedia PDF Downloads 266

Authors: Mahtab Davari, Charles Edward Okon, Somayeh Aghanavesi

Abstract:

Every institute of learning is usually interested in the performance of enrolled students. The level of these performances determines the approach an institute of study may adopt in rendering academic services. The focus of this paper is to evaluate students' academic performance in given courses of study using machine learning methods. This study evaluated various supervised machine learning classification algorithms such as Logistic Regression (LR), Support Vector Machine, Random Forest, Decision Tree, K-Nearest Neighbors, Linear Discriminant Analysis, and Quadratic Discriminant Analysis, using selected features to predict study performance. The accuracy, precision, recall, and F1 score obtained from a 5-Fold Cross-Validation were used to determine the best classification algorithm to predict students’ performances. SVM (using a linear kernel), LDA, and LR were identified as the best-performing machine learning methods. Also, using the LR model, this study identified students' educational habits such as reading and paying attention in class as strong determinants for a student to have an above-average performance. Other important features include the academic history of the student and work. Demographic factors such as age, gender, high school graduation, etc., had no significant effect on a student's performance.

Keywords: student performance, supervised machine learning, classification, cross-validation, prediction

Procedia PDF Downloads 94

Authors: K. I. M. Guerra, L. A. P. Silva, J. C. Leal

Abstract:

This study aims to understand with more amplitude and clarity the behavior of a porous medium where a pressure wave travels, translated into relative displacements inside the material, using mathematical tools derived from topology and symplectic geometry. The paper starts with a given partial differential equation based on the continuity and conservation theorems to describe the traveling wave through the porous body. A solution for this equation is proposed after all boundary, and initial conditions are fixed, and it’s accepted that the solution lies in a manifold U of purely spatial dimensions and that is embedded in the Real n-dimensional manifold, with spatial and kinetic dimensions. It’s shown that the U manifold of lower dimensions than IRna, where it is embedded, inherits properties of the vector spaces existing inside the topology it lies on. Then, a second manifold (U*), embedded in another space called IRnb of stress dimensions, is proposed and there’s a non-degenerative function that maps it into the U manifold. This relation is proved as a transformation in between two corresponding admissible solutions of the differential equation in distinct dimensions and properties, leading to a more visual and intuitive understanding of the whole dynamic process of a stress wave through a porous medium and also highlighting the dimensional invariance of Terzaghi’s theory for any coordinate system.

Keywords: poremechanics, soil dynamics, symplectic geometry, wave propagation

Procedia PDF Downloads 267

Authors: Khanh To Duc, Monica Chiogna, Gianfranco Adimari

Abstract:

With three diagnostic categories, assessment of the performance of diagnostic tests is achieved by the analysis of the receiver operating characteristic (ROC) surface, which generalizes the ROC curve for binary diagnostic outcomes. The volume under the ROC surface (VUS) is a summary index usually employed for measuring the overall diagnostic accuracy. When the true disease status can be exactly assessed by means of a gold standard (GS) test, unbiased nonparametric estimators of the ROC surface and VUS are easily obtained. In practice, unfortunately, disease status verification via the GS test could be unavailable for all study subjects, due to the expensiveness or invasiveness of the GS test. Thus, often only a subset of patients undergoes disease verification. Statistical evaluations of diagnostic accuracy based only on data from subjects with verified disease status are typically biased. This bias is known as verification bias. Here, we consider the problem of correcting for verification bias when continuous diagnostic tests for three-class disease status are considered. We assume that selection for disease verification does not depend on disease status, given test results and other observed covariates, i.e., we assume that the true disease status, when missing, is missing at random. Under this assumption, we discuss several solutions for ROC surface analysis based on imputation and re-weighting methods. In particular, verification bias-corrected estimators of the ROC surface and of VUS are proposed, namely, full imputation, mean score imputation, inverse probability weighting and semiparametric efficient estimators. Consistency and asymptotic normality of the proposed estimators are established, and their finite sample behavior is investigated by means of Monte Carlo simulation studies. Two illustrations using real datasets are also given.

Keywords: imputation, missing at random, inverse probability weighting, ROC surface analysis

Procedia PDF Downloads 391

Authors: Apirak Sombat, Teerapol Saleewong, Poom Kumam, Parin Chaipunya, Wiyada Kumam, Anantachai Padcharoen, Yeol Je Cho, Thana Sutthibutpong

Abstract:

This research is aimed to study a two-step iteration process defined over a finite family of σ-asymptotically quasi-nonexpansive nonself-mappings. The strong convergence is guaranteed under the framework of Banach spaces with some additional structural properties including strict and uniform convexity, reflexivity, and smoothness assumptions. With similar projection technique for nonself-mapping in Hilbert spaces, we hereby use the generalized projection to construct a point within the corresponding domain. Moreover, we have to introduce the use of duality mapping and its inverse to overcome the unavailability of duality representation that is exploit by Hilbert space theorists. We then apply our results for σ-asymptotically quasi-nonexpansive nonself-mappings to solve for ideal efficiency of vector optimization problems composed of finitely many objective functions. We also showed that the obtained solution from our process is the closest to the origin. Moreover, we also give an illustrative numerical example to support our results.

Keywords: asymptotically quasi-nonexpansive nonself-mapping, strong convergence, fixed point, uniformly convex and uniformly smooth Banach space

Procedia PDF Downloads 229

Authors: Gheida J. Shahrour, Martin J. Russell

Abstract:

The 3D body movement signals captured during human-human conversation include clues not only to the content of people’s communication but also to their culture and personality. This paper is concerned with automatic extraction of this information from body movement signals. For the purpose of this research, we collected a novel corpus from 27 subjects, arranged them into groups according to their culture. We arranged each group into pairs and each pair communicated with each other about different topics. A state-of-art recognition system is applied to the problems of person, culture, and topic recognition. We borrowed modeling, classification, and normalization techniques from speech recognition. We used Gaussian Mixture Modeling (GMM) as the main technique for building our three systems, obtaining 77.78%, 55.47%, and 39.06% from the person, culture, and topic recognition systems respectively. In addition, we combined the above GMM systems with Support Vector Machines (SVM) to obtain 85.42%, 62.50%, and 40.63% accuracy for person, culture, and topic recognition respectively. Although direct comparison among these three recognition systems is difficult, it seems that our person recognition system performs best for both GMM and GMM-SVM, suggesting that inter-subject differences (i.e. subject’s personality traits) are a major source of variation. When removing these traits from culture and topic recognition systems using the Nuisance Attribute Projection (NAP) and the Intersession Variability Compensation (ISVC) techniques, we obtained 73.44% and 46.09% accuracy from culture and topic recognition systems respectively.

Keywords: person recognition, topic recognition, culture recognition, 3D body movement signals, variability compensation

Procedia PDF Downloads 516

Authors: Sonali Agarwal

Abstract:

The investor perception about investment avenues is affected to a great degree by the current happenings, within the country, and on the global stage. The influencing events can range from government policies, bilateral trade agreements, election agendas, to changing exchange rates, appreciation and depreciation of currency, recessions, meltdowns, bankruptcies etc. The current research attempts to discover and unravel the effect of various macroeconomic variables (crude oil price, gold price, silver price and USD exchange rate) on the Indian mutual fund industry in general and the chosen funds (Axis Gold Fund, BSL Gold Fund, Kotak Gold Fund & SBI gold fund) in particular. Cointegration tests and Vector error correction equations prove that the chosen variables have strong effect on the NAVs (net asset values) of the mutual funds. However, the greatest influence is felt from the fund’s own past and current information and it is found that when an innovation of fund’s own lagged NAVs is given, variance caused is high that changes the current NAVs markedly. The study helps to highlight the interplay of macroeconomic variables and their repercussion on mutual fund industry.

Keywords: cointegration, Granger causality, impulse response, macroeconomic variables, mutual funds, stationarity, unit root test, variance decomposition, VECM

Procedia PDF Downloads 219

Authors: Assem M. F. Sallam, Ah. El-S. Makled

Abstract:

This paper presents a study on the trajectory of a two stage launch vehicle. The study includes dynamic responses of motion parameters as well as the variation of angles affecting the orientation of the launch vehicle (LV). LV dynamic characteristics including state vector variation with corresponding altitude and velocity for the different LV stages separation, as well as the angle of attack and flight path angles are also discussed. A flight trajectory study for the drop zone of first stage and the jettisoning of fairing are introduced in the mathematical modeling to study their effect. To increase the accuracy of the LV model, atmospheric model is used taking into consideration geographical location and the values of solar flux related to the date and time of launch, accurate atmospheric model leads to enhancement of the calculation of Mach number, which affects the drag force over the LV. The mathematical model is implemented on MATLAB based software (Simulink). The real available experimental data are compared with results obtained from the theoretical computation model. The comparison shows good agreement, which proves the validity of the developed simulation model; the maximum error noticed was generally less than 10%, which is a result that can lead to future works and enhancement to decrease this level of error.

Keywords: launch vehicle modeling, launch vehicle trajectory, mathematical modeling, Matlab- Simulink

Procedia PDF Downloads 257

Authors: Mouldi Djelassi, Jamel Boukhatem

Abstract:

Background: A relation has been established between Islamic banks' activities and interest rates. The aim of this study was to explore the impact of interest rates on the deposits and loans held by Islamic and conventional banks in Saudi Arabia. Methods: A time series data was performed over the period 2008Q1-2020Q2 on eight conventional banks and four Islamic banks. The impacts of interest rate shocks on deposits and loans were identified through panel vector autoregressive models. Results: Impulse response function analysis showed that increasing interest rates reduce loans and conventional deposits. For Islamic banks, deposits are more affected by interest rates than lending. Variance decomposition analysis revealed that deposits contribute to 61% of the Islamic financing variation and only 25% of the conventional loans. Conclusion: Interest rates impacted Islamic banks especially through deposits, which is inconsistent with the theoretical framework. Islamic deposits played an important role in Islamic financing variation and may provide to be a channel for the transmission of the monetary policy in a dual banking system. Monetary policy in Saudi Arabia works in part through “credits” (conventional bank credits) as well as through “money” (conventional and Islamic bank deposits).

Keywords: Islamic banking, interest rates, monetary policy transmission, panel VAR

Procedia PDF Downloads 79

Authors: Zerubabel Mihret

Abstract:

Background: Work-related stress is a pattern of reactions to work demands unmatched by worker’s knowledge, skills, or abilities. Healthcare institutions are considered high-risk and intensive work areas for work-related stress. However, there is the nonexistence of clear and strong data about the magnitude of work-related stress on sanitation workers in hospitals in Ethiopia. The aim of this study was to determine the magnitude of work-related stress among sanitation workers in public hospitals during COVID-19 in Addis Ababa, Ethiopia. Methods: Institution-based cross-sectional study was conducted from October 2021 to February 2022 among 494 sanitation workers who were selected from 4 hospitals. HSE (Health and Safety Executive of UK) standard data collection tool was used, and an interviewer-administered questionnaire was used to collect the data using KOBO collect application. The collected data were cleaned and analyzed using SPSS version 20.0. Both binary and multivariable logistic regression analyses were done to identify important factors having an association with work-related stress. Variables with p-value ≤ 0.25 in the bivariate analysis were entered into the multivariable logistic regression model. A statistically significant level was declared at a p-value ≤ 0.05. Results: This study revealed that the magnitude of work-related stress among sanitation workers was 49.2% (95% CI 45-54). Significant proportions (72.7%) of sanitation workers were dissatisfied with their current job. Sex, age, experience, and chewing khat were significantly associated with work-related stress. Conclusion: Work-related stress is significantly high among sanitation workers. Sex, age, experience, and chewing khat were identified as factors associated with work-related stress. Intervention program focusing on the prevention and control of stress is desired by hospitals.

Keywords: work-related stress, sanitation workers, Likert scale, public hospitals, Ethiopia

Procedia PDF Downloads 52

Authors: Farhad Alizadeh, Alireza Faregh Gharamaleki, Mojtaba Jalilzadeh, Houshang Gholami, Ali Akhoundzadeh

Abstract:

This study presented hybrid pre-processing approach along with a conceptual model to enhance the accuracy of river discharge prediction. In order to achieve this goal, Ensemble Empirical Mode Decomposition algorithm (EEMD), Discrete Wavelet Transform (DWT) and Mutual Information (MI) were employed as a hybrid pre-processing approach conjugated to Least Square Support Vector Machine (LSSVM). A conceptual strategy namely multi-station model was developed to forecast the Souris River discharge more accurately. The strategy used herein was capable of covering uncertainties and complexities of river discharge modeling. DWT and EEMD was coupled, and the feature selection was performed for decomposed sub-series using MI to be employed in multi-station model. In the proposed feature selection method, some useless sub-series were omitted to achieve better performance. Results approved efficiency of the proposed DWT-EEMD-MI approach to improve accuracy of multi-station modeling strategies.

Keywords: river stage-discharge process, LSSVM, discrete wavelet transform, Ensemble Empirical Decomposition Mode, multi-station modeling

Procedia PDF Downloads 151

Authors: Ning Li, Yaner Yan, Yajun Qiao, Shu-qing An

Abstract:

Although seed dispersal is considered to be a key process determining the spatial structure and spread of invasive plant populations, few studies have explicitly addressed the link between dispersal vector behaviour, and seedling recruitment to gain insight into the process of exotic species invasion within a urban landscape. The present study tests the effects of native bird species on the dispersal and recruitment of invasive Phytolacca Americana in an urban garden. We found the invasive population of American pokeweed attracted both generalist species and specialist species to forage and disperse its seeds, with generalists Pycnonotus sinensis and Urocissa erythrorhyncha being by far the most important dispersers. Seedling numbers of P. Americana was strongly affected by perching behavior of bird dispersers. Moreover, two main disperser species, P. sinensis and U. erythrorhyncha govern a high quality dispersal service for P. Americana. Our results highlight the ability of invasive P. americana to recruit seed dispersal agents in urban habitats. However, if the newly recruited species could use the seedling safe site for perching shelter, the invasive plants will get a high regenerate rate in the invasive new habitats thus enhancing their invasive ability.

Keywords: frugivorous birds, phytolacca americana, seed dispersal, urban landscape

Procedia PDF Downloads 520

Authors: Pritika Ramharack, Mahmoud E. S. Soliman

Abstract:

The re-emerging Zika virus is an arthropod-borne virus that has been described to have explosive potential as a worldwide pandemic. The initial transmission of the virus was through a mosquito vector, however, evolving modes of transmission has allowed the spread of the disease over continents. The virus already been linked to irreversible chronic central nervous system (CNS) conditions. The concerns of the scientific and clinical community are the consequences of Zika viral mutations, thus suggesting the urgent need for viral inhibitors. There have been large strides in vaccine development against the virus but there are still no FDA-approved drugs available. Rapid rational drug design and discovery research is fundamental in the production of potent inhibitors against the virus that will not just mask the virus, but destroy it completely. In silico drug design allows for this prompt screening of potential leads, thus decreasing the consumption of precious time and resources. This study demonstrates an optimized and proven screening technique in the discovery of two potential small molecule inhibitors of Zika virus Methyltransferase and RNA-dependent RNA polymerase. This in silico “per-residue energy decomposition pharmacophore” virtual screening approach will be critical in aiding scientists in the discovery of not only effective inhibitors of Zika viral targets, but also a wide range of anti-viral agents.

Keywords: NS5 protein inhibitors, per-residue decomposition, pharmacophore model, virtual screening, Zika virus

Procedia PDF Downloads 201

Authors: Bowen Ding, Yihao Kuang

Abstract:

In recent years, large-scale pre-trained language models have achieved state-of-the-art performance on a variety of natural language processing tasks. The word vectors produced by these language models can be viewed as dense encoded presentations of natural language that in text form. However, it is unknown how much linguistic information is encoded and how. In this paper, we construct several corresponding probing tasks for multiple linguistic information to clarify the encoding capabilities of different language models and performed a visual display. We firstly obtain word presentations in vector form from different language models, including BERT, ELMo, RoBERTa and GPT. Classifiers with a small scale of parameters and unsupervised tasks are then applied on these word vectors to discriminate their capability to encode corresponding linguistic information. The constructed probe tasks contain both semantic and syntactic aspects. The semantic aspect includes the ability of the model to understand semantic entities such as numbers, time, and characters, and the grammatical aspect includes the ability of the language model to understand grammatical structures such as dependency relationships and reference relationships. We also compare encoding capabilities of different layers in the same language model to infer how linguistic information is encoded in the model.

Keywords: language models, probing task, text presentation, linguistic information

Procedia PDF Downloads 78

Authors: Haruna Modibbo Usman, Mustapha Muktar, Nasiru Inuwa

Abstract:

This paper examined the long run relationship between health outcomes and economic growth in Nigeria from 1961 to 2012. Using annual time series data, Augmented Dickey-Fuller (ADF) test is conducted to check the stochastic properties of the variables. Also, the long run relationship among the variables is confirmed based on Johansen Multivariate Cointegration approach whereas the long run and short run dynamics are observed using Vector Error Correction Mechanism (VECM). In addition, VEC Granger causality test is employed to examine the direction of causality among the variables. On the whole, the results obtained revealed the existence of a long run relationship between health outcomes and economic growth in Nigeria and that both life expectancy and crude death rate as measures of health are found to have a long run negative and statistically significant impact on the economic growth over the study period. This is further buttressed by the results of Granger causality test which indicated the existence of unidirectional causality running from life expectancy and crude death rate to economic growth. The study therefore, calls for governments at various levels to create preconditions for health improvements in Nigeria in order to boost the level of health outcomes.

Keywords: cointegration, economic growth, Granger causality, health outcomes, VECM

Procedia PDF Downloads 461

Authors: Cathal Ffrench, Ryan Barrett, Mike Quayle

Abstract:

It is accepted in social psychology that categorization leads to ingroup favoritism, without further thought given to the processes that may co-occur or even precede categorization. These categorizations move away from the conceptualization of the self as a unique social being toward an increasingly collective identity. Subsequently, many individuals derive much of their self-evaluations from these collective identities. The seminal literature on this topic argues that it is primarily categorization that evokes instances of ingroup favoritism. Apropos to these theories, we argue that categorization acts to enhance and further intergroup processes rather than defining them. More accurately, we propose categorization aids initial ingroup contact and this first contact is predictive of subsequent favoritism on individual and collective levels. This analysis focuses on Virtual Interaction APPLication (VIAPPL) based studies, a software interface that builds on the flaws of the original minimal group studies. The VIAPPL allows the exchange of tokens in an intra and inter-group manner. This token exchange is how we classified the first contact. The study involves binary longitudinal analysis to better understand the subsequent exchanges of individuals based on who they first interacted with. Studies were selected on the criteria of evidence of explicit first interactions and two-group designs. Our findings paint a compelling picture in support of a motivated contact hypothesis, which suggests that an individual’s first motivated contact toward another has strong predictive capabilities for future behavior. This contact can lead to habit formation and specific favoritism towards individuals where contact has been established. This has important implications for understanding how group conflict occurs, and how intra-group individual bias can develop.

Keywords: categorization, group dynamics, initial contact, minimal social networks, momentary contact

Procedia PDF Downloads 124

Authors: Mesfin Nigussie

Abstract:

The objectives of the study were to identify factors affecting rural households’ willingness to install biogas plant and amount willingness to pay in order to examine determinants of effective demand for biogas technology. A multistage sampling technique was employed to select 120 respondents for the study. The binary probit regression model was employed to identify factors affecting rural households’ decision to install biogas technology. The probit model result revealed that household size, total household income, access to extension services related to biogas, access to credit service, proximity to water sources, perception of households about the quality of biogas, perception index about attributes of biogas, perception of households about installation cost of biogas and availability of energy source were statistically significant in determining household’s decision to install biogas. Tobit model was employed to examine determinants of rural household’s amount of willingness to pay. Based on the model result, age of the household head, total annual income of the household, access to extension service and availability of other energy source were significant variables that influence willingness to pay. Providing due considerations for extension services, availability of credit or subsidy, improving the quality of biogas technology design and minimizing cost of installation by using locally available materials are the main suggestions of this research that help to create effective demand for biogas technology.

Keywords: biogas technology, effective demand, probit model, tobit model, willingnes to pay

Procedia PDF Downloads 118

Authors: Wei Zhang, Yan He, Yan Wang, Yufeng Li, Chuanpeng Hao

Abstract:

Surface roughness is an important index for evaluating surface quality, needs to be accurately measured to ensure the performance of the workpiece. The roughness measurement based on machine vision involves various image features, some of which are redundant. These redundant features affect the accuracy and speed of the visual approach. Previous research used correlation analysis methods to select the appropriate features. However, this feature analysis is independent and cannot fully utilize the information of data. Besides, blindly reducing features lose a lot of useful information, resulting in unreliable results. Therefore, the focus of this paper is on providing a redundant feature removal approach for visual roughness measurement. In this paper, the statistical methods and gray-level co-occurrence matrix(GLCM) are employed to extract the texture features of machined images effectively. Then, the principal component analysis(PCA) is used to fuse all extracted features into a new one, which reduces the feature dimension and maintains the integrity of the original information. Finally, the relationship between new features and roughness is established by the support vector machine(SVM). The experimental results show that the approach can effectively solve multi-feature information redundancy of machined surface images and provides a new idea for the visual evaluation of surface roughness.

Keywords: feature analysis, machine vision, PCA, surface roughness, SVM

Procedia PDF Downloads 187

Authors: Yavuz Unal, Kemal Polat, H. Erdinc Kocer

Abstract:

In this study, a decision support system comprising three stages has been proposed to detect the disc abnormalities of the lumbar region. In the first stage named the feature extraction, T2-weighted sagittal and axial Magnetic Resonance Images (MRI) were taken from 55 people and then 27 appearance and shape features were acquired from both sagittal and transverse images. In the second stage named the feature weighting process, k-means clustering based feature weighting (KMCBFW) proposed by Gunes et al. Finally, in the third stage named the classification process, the classifier algorithms including multi-layer perceptron (MLP- neural network), support vector machine (SVM), Naïve Bayes, and decision tree have been used to classify whether the subject has lumbar disc or not. In order to test the performance of the proposed method, the classification accuracy (%), sensitivity, specificity, precision, recall, f-measure, kappa value, and computation times have been used. The best hybrid model is the combination of k-means clustering based feature weighting and decision tree in the detecting of lumbar disc disease based on both sagittal and axial MR images.

Keywords: lumbar disc abnormality, lumbar MRI, lumbar spine, hybrid models, hybrid features, k-means clustering based feature weighting

Procedia PDF Downloads 499

Authors: Talal Alshammari, Nasser Alshammari, Mohamed Sedky, Chris Howard

Abstract:

With the widespread adoption of the Internet-connected devices, and with the prevalence of the Internet of Things (IoT) applications, there is an increased interest in machine learning techniques that can provide useful and interesting services in the smart home domain. The areas that machine learning techniques can help advance are varied and ever-evolving. Classifying smart home inhabitants’ Activities of Daily Living (ADLs), is one prominent example. The ability of machine learning technique to find meaningful spatio-temporal relations of high-dimensional data is an important requirement as well. This paper presents a comparative evaluation of state-of-the-art machine learning techniques to classify ADLs in the smart home domain. Forty-two synthetic datasets and two real-world datasets with multiple inhabitants are used to evaluate and compare the performance of the identified machine learning techniques. Our results show significant performance differences between the evaluated techniques. Such as AdaBoost, Cortical Learning Algorithm (CLA), Decision Trees, Hidden Markov Model (HMM), Multi-layer Perceptron (MLP), Structured Perceptron and Support Vector Machines (SVM). Overall, neural network based techniques have shown superiority over the other tested techniques.

Keywords: activities of daily living, classification, internet of things, machine learning, prediction, smart home

Procedia PDF Downloads 321

Authors: Debarati Bhaumik, Diptish Dey

Abstract:

Multilevel models, increasingly deployed in industries such as insurance, food production, and entertainment within functions such as marketing and supply chain management, need to be transparent and ethical. Applications usually result in binary classification within groups or hierarchies based on a set of input features. Using open-source datasets, we demonstrate that popular explainability methods, such as SHAP and LIME, consistently underperform inaccuracy when interpreting these models. They fail to predict the order of feature importance, the magnitudes, and occasionally even the nature of the feature contribution (negative versus positive contribution to the outcome). Besides accuracy, the computational intractability of SHAP for binomial classification is a cause of concern. For transparent and ethical applications of these hierarchical statistical models, sound audit frameworks need to be developed. In this paper, we propose an audit framework for technical assessment of multilevel regression models focusing on three aspects: (i) model assumptions & statistical properties, (ii) model transparency using different explainability methods, and (iii) discrimination assessment. To this end, we undertake a quantitative approach and compare intrinsic model methods with SHAP and LIME. The framework comprises a shortlist of KPIs, such as PoCE (Percentage of Correct Explanations) and MDG (Mean Discriminatory Gap) per feature, for each of these three aspects. A traffic light risk assessment method is furthermore coupled to these KPIs. The audit framework will assist regulatory bodies in performing conformity assessments of AI systems using multilevel binomial classification models at businesses. It will also benefit businesses deploying multilevel models to be future-proof and aligned with the European Commission’s proposed Regulation on Artificial Intelligence.

Keywords: audit, multilevel model, model transparency, model explainability, discrimination, ethics

Procedia PDF Downloads 67

Authors: Hanene Sahli, Aymen Mouelhi, Marwa Hajji, Amine Ben Slama, Mounir Sayadi, Farhat Fnaiech, Radhwane Rachdi

Abstract:

Fetal abnormality is still a public health problem of interest to both mother and baby. Head defect is one of the most high-risk fetal deformities. Fetal head categorization is a sensitive task that needs a massive attention from neurological experts. In this sense, biometrical measurements can be extracted by gynecologist doctors and compared with ground truth charts to identify normal or abnormal growth. The fetal head biometric measurements such as Biparietal Diameter (BPD), Occipito-Frontal Diameter (OFD) and Head Circumference (HC) needs to be monitored, and expert should carry out its manual delineations. This work proposes a new approach to automatically compute BPD, OFD and HC based on morphological characteristics extracted from head shape. Hence, the studied data selected at the same Gestational Age (GA) from the fetal Ultrasound images (US) are classified into two categories: Normal and abnormal. The abnormal subjects include hydrocephalus, microcephaly and dolichocephaly anomalies. By the use of a support vector machines (SVM) method, this study achieved high classification for automated detection of anomalies. The proposed method is promising although it doesn't need expert interventions.

Keywords: biometric measurements, fetal head malformations, machine learning methods, US images

Procedia PDF Downloads 265

Authors: Andras Dezső, Peter Baumli, George Kaptay

Abstract:

The phosphorous is the important components in the steels, because it makes the changing of the mechanical properties and possibly modifying the structure. The phosphorous can be create the Fe3P compounds, what is segregated in the ferrite grain boundary in the intervals of the nano-, or microscale. This intermetallic compound is decreasing the mechanical properties, for example it makes the blue brittleness which means that the brittle created by the segregated particles at 200 ... 300°C. This work describes the phosphide solubility by the other components effect. We make calculations for the Ni, Mo, Cu, S, V, C, Si, Mn, and the Cr elements by the Thermo-Calc software. We predict the effects by approximate functions. The binary Fe-P system has a solubility line, which has a determinating equation. The result is below: lnwo = -3,439 – 1.903/T where the w0 means the weight percent of the maximum soluted concentration of the phosphorous, and the T is the temperature in Kelvin. The equation show that the P more soluble element when the temperature increasing. The nickel, molybdenum, vanadium, silicon, manganese, and the chromium make dependence to the maximum soluted concentration. These functions are more dependent by the elements concentration, which are lower when we put these elements in our steels. The copper, sulphur and carbon do not make effect to the phosphorous solubility. We predict that all of cases the maximum solubility concentration increases when the temperature more and more high. Between 473K and 673 K, in the phase diagram, these systems contain mostly two or three phase eutectoid, and the singe phase, ferritic intervals. In the eutectoid areas the ferrite, the iron-phosphide, and the metal (III)-phospide are in the equilibrium. In these modelling we predicted that which elements are good for avoid the phosphide segregation or not. These datas are important when we make or choose the steels, where the phosphide segregation stopping our possibilities.

Keywords: phosphorous, steel, segregation, thermo-calc software

Procedia PDF Downloads 603

Authors: O. V. Vyazovska

Abstract:

Behavioral efficiency is one of the main principles to be successful in nature. Accuracy of visual discrimination is determined by the attention, learning experience, and memory. In the experimental condition, pigeons’ responses to visual stimuli presented on the screen of the monitor are behaviorally manifested by pecking or not pecking the stimulus, by the number of pecking, reaction time, etc. The higher the probability of rewarding is, the more likely pigeons will respond to the stimulus. We trained 8 pigeons (Columba livia) on a stagewise go/no-go visual discrimination task.16 visual stimuli were created from all possible combinations of four binary dimensions: brightness (dark/bright), size (large/small), line orientation (vertical/horizontal), and shape (circle/square). In the first stage, we presented S+ and 4 S-stimuli: the first that differed in all 4-dimensional values from S+, the second with brightness dimension sharing with S+, the third sharing brightness and orientation with S+, the fourth sharing brightness, orientation and size. Then all 16 stimuli were added. Pigeons rejected correctly 6-8 of 11 new added S-stimuli at the beginning of the second stage. The results revealed that pigeons’ behavior at the beginning of the second stage was controlled by probabilities of rewarding for 4 dimensions learned in the first stage. More or fewer mistakes with dimension discrimination at the beginning of the second stage depended on the number S- stimuli sharing the dimension with S+ in the first stage. A significant inverse correlation between the number of S- stimuli sharing dimension values with S+ in the first stage and the dimensional learning rate at the beginning of the second stage was found. Pigeons were more confident in discrimination of shape and size dimensions. They made mistakes at the beginning of the second stage, which were not associated with these dimensions. Thus, the received results help elucidate the principles of dimensional stimulus control during learning compound multidimensional visual stimuli.

Keywords: visual go/no go discrimination, selective attention, dimensional stimulus control, pigeon

Procedia PDF Downloads 117

Authors: Yahaya Abdullahi, Ben Coetzee, Linda Van Den Berg

Abstract:

The study aimed at establishing the global positioning system (GPS) determined singles match characteristics that act as predictors of successful and less-successful male singles badminton players’ group classification. Twenty-two (22) male single players (aged: 23.39 ± 3.92 years; body stature: 177.11 ± 3.06cm; body mass: 83.46 ± 14.59kg) who represented 10 African countries participated in the study. Players were categorised as successful and less-successful players according to the results of five championships’ of the 2014/2015 season. GPS units (MinimaxX V4.0), Polar Heart Rate Transmitter Belts and digital video cameras were used to collect match data. GPS-related variables were corrected for match duration and independent t-tests, a cluster analysis and a binary forward stepwise logistic regression were calculated. A Receiver Operating Characteristic Curve (ROC) was used to determine the validity of the group classification model. High-intensity accelerations per second were identified as the only GPS-determined variable that showed a significant difference between groups. Furthermore, only high-intensity accelerations per second (p=0.03) and low-intensity efforts per second (p=0.04) were identified as significant predictors of group classification with 76.88% of players that could be classified back into their original groups by making use of the GPS-based logistic regression formula. The ROC showed a value of 0.87. The identification of the last-mentioned GPS-related variables for the attainment of badminton performances, emphasizes the importance of using badminton drills and conditioning techniques to not only improve players’ physical fitness levels but also their abilities to accelerate at high intensities.

Keywords: badminton, global positioning system, match analysis, inertial movement analysis, intensity, effort

Procedia PDF Downloads 169

Authors: Bo Hu, Shibo Wang, Haoyang Zhang, Jie Shi

Abstract:

With the development of sonar detection technology, the acoustic stealth technology of underwater vehicles is facing severe challenges. The underwater acoustic coating is developing towards the direction of low-frequency absorption capability and broad absorption frequency bandwidth. In this paper, an acoustic model of underwater acoustic coating of composite material embedded with periodical steel structure is presented. The model has multiple high absorption peaks in the frequency range of 1kHz-8kHz, where achieves high sound absorption and broad bandwidth performance. It is found that the frequencies of the absorption peaks are related to the classic half-wavelength transmission principle. The sound absorption performance of the acoustic model is investigated by the finite element method using COMSOL software. The sound absorption mechanism of the proposed model is explained by the distributions of the displacement vector field. The influence of geometric parameters of periodical steel structure, including thickness and distance, on the sound absorption ability of the proposed model are further discussed. The acoustic model proposed in this study provides an idea for the design of underwater low-frequency broadband acoustic coating, and the results shows the possibility and feasibility for practical underwater application.

Keywords: acoustic coating, composite material, broad frequency bandwidth, sound absorption performance

Procedia PDF Downloads 147

Authors: R. R. Ramsheeja, R. Sreeraj

Abstract:

For view the internal structures of the human body such as liver, brain, kidney etc have a wide range of different modalities for medical images are provided nowadays. Computer Tomography is one of the most significant medical image modalities. In this paper use CT liver images for study the use of automatic computer aided techniques to calculate the volume of the liver tumor. Segmentation method is used for the detection of tumor from the CT scan is proposed. Gaussian filter is used for denoising the liver image and Adaptive Thresholding algorithm is used for segmentation. Multiple Region Of Interest(ROI) based method that may help to characteristic the feature different. It provides a significant impact on classification performance. Due to the characteristic of liver tumor lesion, inherent difficulties appear selective. For a better performance, a novel proposed system is introduced. Multiple ROI based feature selection and classification are performed. In order to obtain of relevant features for Support Vector Machine(SVM) classifier is important for better generalization performance. The proposed system helps to improve the better classification performance, reason in which we can see a significant reduction of features is used. The diagnosis of liver cancer from the computer tomography images is very difficult in nature. Early detection of liver tumor is very helpful to save the human life.

Keywords: computed tomography (CT), multiple region of interest(ROI), feature values, segmentation, SVM classification

Procedia PDF Downloads 488

Authors: Zhuoran Shan, Li Wan, Xianchun Zhang

Abstract:

With an increasing trend of regionalization and polycentricity in Chinese contemporary big cities, “sub-region life circle” turns to be an effective method on rational organization of urban function and spatial structure. By the method of questionnaire, network big data, route inversion on internet map, GIS spatial analysis and logistic regression, this article makes research on characteristic and formation mechanism of “sub-region life circle” based on human high-probability daily activity in Chinese big cities. Firstly, it shows that “sub-region life circle” has been a new general spatial sphere of residents' high-probability daily activity and mobility in China. Unlike the former analysis of the whole metropolitan or the micro community, “sub-region life circle” has its own characteristic on geographical sphere, functional element, spatial morphology and land distribution. Secondly, according to the analysis result with Binary Logistic Regression Model, the research also shows that seven factors including land-use mixed degree and bus station density impact the formation of “sub-region life circle” most, and then analyzes the index critical value of each factor. Finally, to establish a smarter “sub-region life circle”, this paper indicates that several strategies including jobs-housing fit, service cohesion and space reconstruction are the keys for its spatial organization optimization. This study expands the further understanding of cities' inner sub-region spatial structure based on human daily activity, and contributes to the theory of “life circle” in urban's meso-scale.

Keywords: sub-region life circle, characteristic, formation mechanism, human activity, spatial structure

Procedia PDF Downloads 274

Authors: Bowei Yuan, Shi Li, Liuyang Song, Huaqing Wang, Lingli Cui

Abstract:

Convolutional neural networks (CNN) have received extensive attention in the field of fault diagnosis. Many fault diagnosis methods use CNN for fault type identification. However, when the amount of raw data collected by sensors is massive, the neural network needs to perform a time-consuming classification task. In this paper, a mechanical fault diagnosis method based on vibration signal down-sampling and the improved one-dimensional convolutional neural network is proposed. Through the robust principal component analysis, the low-rank feature matrix of a large amount of raw data can be separated, and then down-sampling is realized to reduce the subsequent calculation amount. In the improved one-dimensional CNN, a smaller convolution kernel is used to reduce the number of parameters and computational complexity, and regularization is introduced before the fully connected layer to prevent overfitting. In addition, the multi-connected layers can better generalize classification results without cumbersome parameter adjustments. The effectiveness of the method is verified by monitoring the signal of the centrifugal pump test bench, and the average test accuracy is above 98%. When compared with the traditional deep belief network (DBN) and support vector machine (SVM) methods, this method has better performance.

Keywords: fault diagnosis, vibration signal down-sampling, 1D-CNN

Procedia PDF Downloads 106