Search results for: empirical bayesian kriging regression prediction

Authors: Li Kewen, Su Zhaoxin, Wang Xingmou, Zhu Jian Bing

Abstract:

Conventional reservoir prediction methods ar not sufficient to explore the implicit relation between seismic attributes, and thus data utilization is low. In order to improve the predictive classification accuracy of reservoir lithology, this paper proposes a deep learning lithology prediction method based on ResNet (Residual Neural Network) and SENet (Squeeze-and-Excitation Neural Network). The neural network model is built and trained by using seismic attribute data and lithology data of Shengli oilfield, and the nonlinear mapping relationship between seismic attribute and lithology marker is established. The experimental results show that this method can significantly improve the classification effect of reservoir lithology, and the classification accuracy is close to 70%. This study can effectively predict the lithology of undrilled area and provide support for exploration and development.

Keywords: convolutional neural network, lithology, prediction of reservoir, seismic attributes

Procedia PDF Downloads 158

Authors: Inten Meutia, Emylia Yuniarti

Abstract:

The objective of this study is to analyze the effects of interest rate, the rate of return of Islamic banks on the amount of mudarabah deposits in Islamic banks. In analyzing the effect of rate of return in the Islamic banks and interest rate risk in the conventional banks, the 1-month Islamic deposit rate of return and 1 month fixed deposit interest rate of a total Islamic deposit are considered. Using data covering the period from January 2010 to Sepember 2013, the study applies the regression analysis to analyze the effect between variable and independence t-test to analyze the mean difference between rate of return and rate of interest. Regression analysis shows that rate of return have significantly negative influence on mudarabah deposits, while interest rate have negative influence but not significant. The result of independent t test shows that the interest rate is not different from the rate of return in Islamic Bank. It supports the hyphotesis that rate of return in Islamic banking mimic rate of interest in conventional bank. The results of the study have important implications on the risk management practices of the Islamic banks in Indonesia.

Keywords: conventional bank, interest rate, Islamic bank, rate of return

Procedia PDF Downloads 487

Authors: Trupti Diwan

Abstract:

Predicting student failure at school has changed into a difficult challenge due to both the large number of factors that can affect the reduced performance of students and the imbalanced nature of these kinds of data sets. This paper surveys the two elements needed to make prediction on Students’ Academic Performances which are parameters and methods. This paper also proposes a framework for predicting the performance of engineering students. Genetic programming can be used to predict student failure/success. Ranking algorithm is used to rank students according to their credit points. The framework can be used as a basis for the system implementation & prediction of students’ Academic Performance in Higher Learning Institute.

Keywords: classification, educational data mining, student failure, grammar-based genetic programming

Procedia PDF Downloads 403

Authors: Seyed Esmail Seyedi Bariran, Khairul Salleh Mohamed Sahari

Abstract:

In multi-echelon supply chain networks, optimal supplier selection significantly depends on the accuracy of suppliers’ performance prediction. Different methods of multi criteria decision making such as ANN, GA, Fuzzy, AHP, etc have been previously used to predict the supplier performance but the “black-box” characteristic of these methods is yet a major concern to be resolved. Therefore, the primary objective in this paper is to implement an artificial intelligence-based gene expression programming (GEP) model to compare the prediction accuracy with that of ANN. A full factorial design with %95 confidence interval is initially applied to determine the appropriate set of criteria for supplier performance evaluation. A test-train approach is then utilized for the ANN and GEP exclusively. The training results are used to find the optimal network architecture and the testing data will determine the prediction accuracy of each method based on measures of root mean square error (RMSE) and correlation coefficient (R2). The results of a case study conducted in Supplying Automotive Parts Co. (SAPCO) with more than 100 local and foreign supply chain members revealed that, in comparison with ANN, gene expression programming has a significant preference in predicting supplier performance by referring to the respective RMSE and R-squared values. Moreover, using GEP, a mathematical function was also derived to solve the issue of ANN black-box structure in modeling the performance prediction.

Keywords: Supplier Performance Prediction, ANN, GEP, Automotive, SAPCO

Procedia PDF Downloads 400

Authors: Ipek Kivanc, Demet Ozgur-Unluakin

Abstract:

Over the past years, technological innovations and advancements have played an important role in the industrial world. Due to technological improvements, the degree of complexity of the systems has increased. Hence, all systems are getting more uncertain that emerges from increased complexity, resulting in more cost. It is challenging to cope with this situation. So, implementing efficient planning of maintenance activities in such systems are getting more essential. Partially Observable Markov Decision Processes (POMDPs) are powerful tools for stochastic sequential decision problems under uncertainty. Although maintenance optimization in a dynamic environment can be modeled as such a sequential decision problem, POMDPs are not widely used for tackling maintenance problems. However, they can be well-suited frameworks for obtaining optimal maintenance policies. In the classical representation of the POMDP framework, the system is denoted by a single node which has multiple states. The main drawback of this classical approach is that the state space grows exponentially with the number of state variables. On the other side, factored representation of POMDPs enables to simplify the complexity of the states by taking advantage of the factored structure already available in the nature of the problem. The main idea of factored POMDPs is that they can be compactly modeled through dynamic Bayesian networks (DBNs), which are graphical representations for stochastic processes, by exploiting the structure of this representation. This study aims to demonstrate how maintenance planning of dynamic systems can be modeled with factored POMDPs. An empirical maintenance planning problem of a dynamic system consisting of four partially observable components deteriorating in time is designed. To solve the empirical model, we resort to Symbolic Perseus solver which is one of the state-of-the-art factored POMDP solvers enabling approximate solutions. We generate some more predefined policies based on corrective or proactive maintenance strategies. We execute the policies on the empirical problem for many replications and compare their performances under various scenarios. The results show that the computed policies from the POMDP model are superior to the others. Acknowledgment: This work is supported by the Scientific and Technological Research Council of Turkey (TÜBİTAK) under grant no: 117M587.

Keywords: factored representation, maintenance, multi-component system, partially observable Markov decision processes

Procedia PDF Downloads 117

Authors: Hervice Roméo Fogno Fotsoa, Germaine Djuidje Kenmoe, Claude Vidal Aloyem Kazé

Abstract:

One of the main applications of machine learning is the prediction of time series. But a more accurate prediction requires a more optimal model of machine learning. Several optimization techniques have been developed, but without considering the input variables disposition of the system. Thus, this work aims to present a new machine learning architecture optimization technique based on their optimal input variables disposition. The validations are done on the prediction of wind time series, using data collected in Cameroon. The number of possible dispositions with four input variables is determined, i.e., twenty-four. Each of the dispositions is used to perform the prediction, with the main criteria being the training and prediction performances. The results obtained from a static architecture and a dynamic architecture of neural networks have shown that these performances are a function of the input variable's disposition, and this is in a different way from the architectures. This analysis revealed that it is necessary to take into account the input variable's disposition for the development of a more optimal neural network model. Thus, a new neural network training algorithm is proposed by introducing the search for the optimal input variables disposition in the traditional back-propagation algorithm. The results of the application of this new optimization approach on the two single neural network architectures are compared with the previously obtained results step by step. Moreover, this proposed approach is validated in a collaborative optimization method with a single objective optimization technique, i.e., genetic algorithm back-propagation neural networks. From these comparisons, it is concluded that each proposed model outperforms its traditional model in terms of training and prediction performance of time series. Thus the proposed optimization approach can be useful in improving the accuracy of time series forecasts. This proves that the proposed optimization approach can be useful in improving the accuracy of time series prediction based on machine learning.

Keywords: input variable disposition, machine learning, optimization, performance, time series prediction

Procedia PDF Downloads 76

Authors: Ekele Alih, Enejo Jalija

Abstract:

Female Genital Mutilation (FGM) is made up of three types namely: Clitoridectomy, Excision and Infibulation. In this study, we examine the factors responsible for FGM in order to identify the causality variables in a logistic regression approach. From the result of the survey conducted by the Public Health Division, Nigeria Institute of Medical Research, Yaba, Lagos State, the tau statistic, τ was used to screen 9 factors that causes FGM in order to select few of the predictors before multiple regression equation is obtained. The need for this may be that the sample size may not be able to sustain having a regression with all the predictors or to avoid multi-collinearity. A total of 300 respondents, comprising 150 adult males and 150 adult females were selected for the household survey based on the multi-stage sampling procedure. The tau statistic,

Keywords: female genital mutilation, logistic regression, tau statistic, African society

Procedia PDF Downloads 232

Authors: Belkacem Selma, Boumediene Selma, Tourkia Guerzou, Abbes Labdelli

Abstract:

Natural disasters have occurred and will continue to cause human and material damage. Therefore, the idea of "preventing" natural disasters will never be possible. However, their prediction is possible with the advancement of technology. Even if natural disasters are effectively inevitable, their consequences may be partly controlled. The rapid growth and progress of artificial intelligence (AI) had a major impact on the prediction of natural disasters and risk assessment which are necessary for effective disaster reduction. The Earthquakes prediction to prevent the loss of human lives and even property damage is an important factor; that is why it is crucial to develop techniques for predicting this natural disaster. This present study aims to analyze the ability of artificial neural networks (ANNs) to predict earthquakes that occur in a given area. The used data describe the problem of high energy (higher than 10^4J) seismic bumps forecasting in a coal mine using two long walls as an example. For this purpose, seismic bumps data obtained from mines has been analyzed. The results obtained show that the ANN with high accuracy was able to predict earthquake parameters; the classification accuracy through neural networks is more than 94%, and that the models developed are efficient and robust and depend only weakly on the initial database.

Keywords: earthquake prediction, ANN, seismic bumps

Procedia PDF Downloads 106

Authors: Georgios Charizanos, Haydar Demirhan, Duygu Icen

Abstract:

Two of the most impactful issues in classical logistic regression are class imbalance and complete separation. These can result in model predictions heavily leaning towards the imbalanced class on the binary response variable or over-fitting issues. Fuzzy methodology offers key solutions for handling these problems. However, most studies propose the transformation of the binary responses into a continuous format limited within [0,1]. This is called the possibilistic approach within fuzzy logistic regression. Following this approach is more aligned with straightforward regression since a logit-link function is not utilized, and fuzzy probabilities are not generated. In contrast, we propose a method of fuzzifying binary response variables that allows for the use of the logit-link function; hence, a probabilistic fuzzy logistic regression model with the Monte Carlo method. The fuzzy probabilities are then classified by selecting a fuzzy threshold. Different combinations of fuzzy and crisp input, output, and coefficients are explored, aiming to understand which of these perform better under different conditions of imbalance and separation. We conduct numerical experiments using both synthetic and real datasets to demonstrate the performance of the fuzzy logistic regression framework against seven crisp machine learning methods. The proposed framework shows better performance irrespective of the degree of imbalance and presence of separation in the data, while the considered machine learning methods are significantly impacted.

Keywords: fuzzy logistic regression, fuzzy, logistic, machine learning

Procedia PDF Downloads 47

Authors: Achut Manandhar, Kenneth D. Morton, Peter A. Torrione, Leslie M. Collins

Abstract:

The current trends in affect recognition research are to consider continuous observations from spontaneous natural interactions in people using multiple feature modalities, and to represent affect in terms of continuous dimensions, incorporate spatio-temporal correlation among affect dimensions, and provide fast affect predictions. These research efforts have been propelled by a growing effort to develop affect recognition system that can be implemented to enable seamless real-time human-computer interaction in a wide variety of applications. Motivated by these desired attributes of an affect recognition system, in this work a multi-dimensional affect prediction approach is proposed by integrating multivariate Relevance Vector Machine (MVRVM) with a recently developed Output-associative Relevance Vector Machine (OARVM) approach. The resulting approach can provide fast continuous affect predictions by jointly modeling the multiple affect dimensions and their correlations. Experiments on the RECOLA database show that the proposed approach performs competitively with the OARVM while providing faster predictions during testing.

Keywords: dimensional affect prediction, output-associative RVM, multivariate regression, fast testing

Procedia PDF Downloads 263

Authors: Yong Zhao, Jian He, Cheng Zhang

Abstract:

Cardiovascular diseases caused by hypertension are extremely threatening to human health, and early diagnosis of hypertension can save a large number of lives. Traditional hypertension detection methods require special equipment and are difficult to detect continuous blood pressure changes. In this regard, this paper first analyzes the principle of heart rate variability (HRV) and introduces sliding window and power spectral density (PSD) to analyze the time domain features and frequency domain features of HRV, and secondly, designs an HRV-based hypertension prediction network by combining Resnet, attention mechanism, and multilayer perceptron, which extracts the frequency domain through the improved ResNet18 features through a modified ResNet18, its fusion with time-domain features through an attention mechanism, and the auxiliary prediction of hypertension through a multilayer perceptron. Finally, the network was trained and tested using the publicly available SHAREE dataset on PhysioNet, and the test results showed that this network achieved 92.06% prediction accuracy for hypertension and outperformed K Near Neighbor(KNN), Bayes, Logistic, and traditional Convolutional Neural Network(CNN) models in prediction performance.

Keywords: feature extraction, heart rate variability, hypertension, residual networks

Procedia PDF Downloads 73

Authors: Leonardo Juan Ramírez López, Javier Oswaldo Rodriguez Velasquez

Abstract:

We have designed a Holter that measures the heart´s activity for over 24 hours, implemented a prediction methodology, and generate alarms as well as indicators to patients and treating physicians. Various diagnostic advances have been developed in clinical cardiology thanks to Holter implementation; however, their interpretation has largely been conditioned to clinical analysis and measurements adjusted to diverse population characteristics, thus turning it into a subjective examination. This, however, requires vast population studies to be validated that, in turn, have not achieved the ultimate goal: mortality prediction. Given this context, our Insight Research Group developed a mathematical methodology that assesses cardiac dynamics through entropy and probability, creating a numerical and geometrical attractor which allows quantifying the normalcy of chronic and acute disease as well as the evolution between such states, and our Tigum Research Group developed a holter device with 12 channels and advanced computer software. This has been shown in different contexts with 100% sensitivity and specificity results.

Keywords: attractor , cardiac, entropy, holter, mathematical , prediction

Procedia PDF Downloads 149

Authors: S.K. Bhandari, E. Veneklaas, L. McCaw, R. Mazanec, K. Whitford, M. Renton

Abstract:

Eucalyptus marginata, E. diversicolor and Corymbia calophylla form widespread forests in south-west Western Australia (SWWA). These forests have economic and ecological importance, and therefore, tree growth and sustainable management are of high priority. This paper aimed to analyse and model the growth of these species at both stand and individual levels, but this presentation will focus on predicting the growth of E. Marginata at the individual tree level. More specifically, the study wanted to investigate how well individual E. marginata tree growth could be predicted by considering the diameter and height of the tree at the start of the growth period, and whether this prediction could be improved by also accounting for the competition from neighbouring trees in different ways. The study also wanted to investigate how many neighbouring trees or what neighbourhood distance needed to be considered when accounting for competition. To achieve this aim, the Pearson correlation coefficient was examined among competition indices (CIs), between CIs and dbh growth, and selected the competition index that can best predict the diameter growth of individual trees of E. marginata forest managed under different thinning regimes at Inglehope in SWWA. Furthermore, individual tree growth models were developed using simple linear regression, multiple linear regression, and linear mixed effect modelling approaches. Individual tree growth models were developed for thinned and unthinned stand separately. The developed models were validated using two approaches. In the first approach, models were validated using a subset of data that was not used in model fitting. In the second approach, the model of the one growth period was validated with the data of another growth period. Tree size (diameter and height) was a significant predictor of growth. This prediction was improved when the competition was included in the model. The fit statistic (coefficient of determination) of the model ranged from 0.31 to 0.68. The model with spatial competition indices validated as being more accurate than with non-spatial indices. The model prediction can be optimized if 10 to 15 competitors (by number) or competitors within ~10 m (by distance) from the base of the subject tree are included in the model, which can reduce the time and cost of collecting the information about the competitors. As competition from neighbours was a significant predictor with a negative effect on growth, it is recommended including neighbourhood competition when predicting growth and considering thinning treatments to minimize the effect of competition on growth. These model approaches are likely to be useful tools for the conservations and sustainable management of forests of E. marginata in SWWA. As a next step in optimizing the number and distance of competitors, further studies in larger size plots and with a larger number of plots than those used in the present study are recommended.

Keywords: competition, growth, model, thinning

Procedia PDF Downloads 107

Authors: Mohamed Aly, Aliaa Rehan Youssef, Emad Sawerees, Mounir Guirgis

Abstract:

Relevance: Osteoarthritis (OA) is the most common degenerative disease seen in all populations. It causes disability and substantial socioeconomic burden. Evidence supports that exercise are the most effective conservative treatment for patients with OA. Therapists experience and clinical judgment play major role in exercise prescription and scientific evidence for this regard is lacking. The development of clinical prediction rules to identify patients who are most likely benefit from exercise may help solving this dilemma. Purpose: This study investigated whether body mass index and functional ability at baseline can predict patients’ response to a selected exercise program. Approach: Fifty-six patients, aged 35 to 65 years, completed an exercise program consisting of open kinetic chain strengthening and passive stretching exercises. The program was given for 3 sessions per week, 45 minutes per session, for 6 weeks Evaluation: At baseline and post treatment, pain severity was assessed using the numerical pain rating scale, whereas functional ability was being assessed by step test (ST), time up and go test (TUG) and 50 feet time walk test (50 FTW). After completing the program, global rate of change (GROC) score of greater than 4 was used to categorize patients as successful and non-successful. Thirty-eight patients (68%) had successful response to the intervention. Logistic regression showed that BMI and 50 FTW test were the only significant predictors. Based on the results, patients with BMI less than 34.71 kg/m2 and 50 FTW test less than 25.64 sec are 68% to 89% more likely to benefit from the exercise program. Conclusions: Clinicians should consider the described strengthening and flexibility exercise program for patents with BMI less than 34.7 Kg/m2 and 50 FTW faster than 25.6 seconds. The validity of these predictors should be investigated for other exercise.

Keywords: clinical prediction rule, knee osteoarthritis, physical therapy exercises, validity

Procedia PDF Downloads 395

Authors: Ruchika Malhotra, Megha Khanna

Abstract:

The development of change prediction models can help the software practitioners in planning testing and inspection resources at early phases of software development. However, a major challenge faced during the training process of any classification model is the imbalanced nature of the software quality data. A data with very few minority outcome categories leads to inefficient learning process and a classification model developed from the imbalanced data generally does not predict these minority categories correctly. Thus, for a given dataset, a minority of classes may be change prone whereas a majority of classes may be non-change prone. This study explores various alternatives for adeptly handling the imbalanced software quality data using different sampling methods and effective MetaCost learners. The study also analyzes and justifies the use of different performance metrics while dealing with the imbalanced data. In order to empirically validate different alternatives, the study uses change data from three application packages of open-source Android data set and evaluates the performance of six different machine learning techniques. The results of the study indicate extensive improvement in the performance of the classification models when using resampling method and robust performance measures.

Keywords: change proneness, empirical validation, imbalanced learning, machine learning techniques, object-oriented metrics

Procedia PDF Downloads 398

Authors: Meera Tandan, Mohan Timilsina, Martin Cormican, Akke Vellinga

Abstract:

Background: In general practice, many infections are treated empirically without microbiological confirmation. Understanding susceptibility of antimicrobials during empirical prescribing can be helpful to reduce inappropriate prescribing. This study aims to apply a prediction model using a decision tree approach to predict the antimicrobial resistance (AMR) of urinary tract infections (UTI) based on non-clinical features of patients over 65 years. Decision tree models are a novel idea to predict the outcome of AMR at an initial stage. Method: Data was extracted from the database of the microbiological laboratory of the University Hospitals Galway on all antimicrobial susceptibility testing (AST) of urine specimens from patients over the age of 65 from January 2011 to December 2014. The primary endpoint was resistance to common antimicrobials (Nitrofurantoin, trimethoprim, ciprofloxacin, co-amoxiclav and amoxicillin) used to treat UTI. A classification and regression tree (CART) model was generated with the outcome ‘resistant infection’. The importance of each predictor (the number of previous samples, age, gender, location (nursing home, hospital, community) and causative agent) on antimicrobial resistance was estimated. Sensitivity, specificity, negative predictive (NPV) and positive predictive (PPV) values were used to evaluate the performance of the model. Seventy-five percent (75%) of the data were used as a training set and validation of the model was performed with the remaining 25% of the dataset. Results: A total of 9805 UTI patients over 65 years had their urine sample submitted for AST at least once over the four years. E.coli, Klebsiella, Proteus species were the most commonly identified pathogens among the UTI patients without catheter whereas Sertia, Staphylococcus aureus; Enterobacter was common with the catheter. The validated CART model shows slight differences in the sensitivity, specificity, PPV and NPV in between the models with and without the causative organisms. The sensitivity, specificity, PPV and NPV for the model with non-clinical predictors was between 74% and 88% depending on the antimicrobial. Conclusion: The CART models developed using non-clinical predictors have good performance when predicting antimicrobial resistance. These models predict which antimicrobial may be the most appropriate based on non-clinical factors. Other CART models, prospective data collection and validation and an increasing number of non-clinical factors will improve model performance. The presented model provides an alternative approach to decision making on antimicrobial prescribing for UTIs in older patients.

Keywords: antimicrobial resistance, urinary tract infection, prediction, decision tree

Procedia PDF Downloads 234

Authors: Mo-Se Lee, Cheol-Hwi Ahn, Kee-Young Kwahk, Hyunchul Ahn

Abstract:

Over the past decade, deep learning has been in spotlight among various machine learning algorithms. In particular, CNN (Convolutional Neural Network), which is known as effective solution for recognizing and classifying images, has been popularly applied to classification and prediction problems in various fields. In this study, we try to apply CNN to stock market prediction, one of the most challenging tasks in the machine learning research. In specific, we propose to apply CNN as the binary classifier that predicts stock market direction (up or down) by using a graph as its input. That is, our proposal is to build a machine learning algorithm that mimics a person who looks at the graph and predicts whether the trend will go up or down. Our proposed model consists of four steps. In the first step, it divides the dataset into 5 days, 10 days, 15 days, and 20 days. And then, it creates graphs for each interval in step 2. In the next step, CNN classifiers are trained using the graphs generated in the previous step. In step 4, it optimizes the hyper parameters of the trained model by using the validation dataset. To validate our model, we will apply it to the prediction of KOSPI200 for 1,986 days in eight years (from 2009 to 2016). The experimental dataset will include 14 technical indicators such as CCI, Momentum, ROC and daily closing price of KOSPI200 of Korean stock market.

Keywords: convolutional neural network, deep learning, Korean stock market, stock market prediction

Procedia PDF Downloads 410

Authors: Afroze Ibrahim Baqapuri, Ilya Trofimov

Abstract:

Sponsored search is a multi-billion dollar industry and makes up a major source of revenue for search engines (SE). Click-through-rate (CTR) estimation plays a crucial role for ads selection, and greatly affects the SE revenue, advertiser traffic and user experience. We propose a novel architecture of solving CTR prediction problem by combining artificial neural networks (ANN) with decision trees. First, we compare ANN with respect to other popular machine learning models being used for this task. Then we go on to combine ANN with MatrixNet (proprietary implementation of boosted trees) and evaluate the performance of the system as a whole. The results show that our approach provides a significant improvement over existing models.

Keywords: neural networks, sponsored search, web advertisement, click prediction, click-through rate

Procedia PDF Downloads 552

Authors: Akram Khaleghei, Ghosheh Balagh, Viliam Makis

Abstract:

In this paper, we investigate the residual life prediction problem for a partially observable system subject to two failure modes, namely a catastrophic failure and a failure due to the system degradation. The system is subject to condition monitoring and the degradation process is described by a hidden Markov model with unknown parameters. The parameter estimation procedure based on an EM algorithm is developed and the formulas for the conditional reliability function and the mean residual life are derived, illustrated by a numerical example.

Keywords: partially observable system, hidden Markov model, competing risks, residual life prediction

Procedia PDF Downloads 391

Authors: Oluwayemisi O. Alaba, John O. Olaomi

Abstract:

The 2013 Nigerian Demographic Health Survey (NDHS) data was used to investigate the determinants of family size in Nigeria using the geo-additive model. The fixed effect of categorical covariates were modelled using the diffuse prior, P-spline with second-order random walk for the nonlinear effect of continuous variable, spatial effects followed Markov random field priors while the exchangeable normal priors were used for the random effects of the community and household. The Negative Binomial distribution was used to handle overdispersion of the dependent variable. Inference was fully Bayesian approach. Results showed a declining effect of secondary and higher education of mother, Yoruba tribe, Christianity, family planning, mother giving birth by caesarean section and having a partner who has secondary education on family size. Big family size is positively associated with age at first birth, number of daughters in a household, being gainfully employed, married and living with partner, community and household effects.

Keywords: Bayesian analysis, family size, geo-additive model, negative binomial

Procedia PDF Downloads 517

Authors: Shokrya Saleh A. Alshqaq, Abdullah Ali H. Ahmadini

Abstract:

The Schwarz information criterion (SIC) is a popular tool for selecting the best variables in regression datasets. However, SIC is defined using an unbounded estimator, namely, the least-squares (LS), which is highly sensitive to outlying observations, especially bad leverage points. A method for robust variable selection based on SIC for linear regression models is thus needed. This study investigates the robustness properties of SIC by deriving its influence function and proposes a robust SIC based on the MM-estimation scale. The aim of this study is to produce a criterion that can effectively select accurate models in the presence of vertical outliers and high leverage points. The advantages of the proposed robust SIC is demonstrated through a simulation study and an analysis of a real dataset.

Keywords: influence function, robust variable selection, robust regression, Schwarz information criterion

Procedia PDF Downloads 120

Authors: Jayanta Pokharel, Gokarna Aryal, Netra Kanaal, Chris Tsokos

Abstract:

Investment in stocks and shares aims to seek potential gains while weighing the risk of future needs, such as retirement, children's education etc. Analysis of the behavior of the stock market returns and making prediction is important for investors to mitigate risk on investment. Historically, the normal variance models have been used to describe the behavior of stock market returns. However, the returns of the financial assets are actually skewed with higher kurtosis, heavier tails, and a higher center than the normal distribution. The Laplace distribution and its family are natural candidates for modeling stock returns. The Variance-Gamma (VG) distribution is the most sought-after distributions for modeling asset returns and has been extensively discussed in financial literatures. In this paper, it explore the other Laplace family, such as Asymmetric Laplace, Skewed Laplace, Kumaraswamy Laplace (KS) together with Variance-Gamma to model the weekly returns of the S&P 500 Index and it's eleven business sector indices. The method of maximum likelihood is employed to estimate the parameters of the distributions and our empirical inquiry shows that the Kumaraswamy Laplace distribution performs much better for stock returns modeling among the choice of distributions used in this study and in practice, KS can be used as a strong alternative to VG distribution.

Keywords: stock returns, variance-gamma, kumaraswamy laplace, maximum likelihood

Procedia PDF Downloads 49

Authors: Tahmidul Islam

Abstract:

Propensity Score Matching (PSM) technique has been widely used for estimating causal effect of treatment in observational studies. One major step of implementing PSM is estimating the propensity score (PS). Logistic regression model with additive linear terms of covariates is most used technique in many studies. Logistics regression model is also used with cubic splines for retaining flexibility in the model. However, choosing the functional form of the logistic regression model has been a question since the effectiveness of PSM depends on how accurately the PS been estimated. In many situations, the linearity assumption of linear logistic regression may not hold and non-linear relation between the logit and the covariates may be appropriate. One can estimate PS using machine learning techniques such as random forest, neural network etc for more accuracy in non-linear situation. In this study, an attempt has been made to compare the efficacy of Generalized Additive Model (GAM) in various linear and non-linear settings and compare its performance with usual logistic regression. GAM is a non-parametric technique where functional form of the covariates can be unspecified and a flexible regression model can be fitted. In this study various simple and complex models have been considered for treatment under several situations (small/large sample, low/high number of treatment units) and examined which method leads to more covariate balance in the matched dataset. It is found that logistic regression model is impressively robust against inclusion quadratic and interaction terms and reduces mean difference in treatment and control set equally efficiently as GAM does. GAM provided no significantly better covariate balance than logistic regression in both simple and complex models. The analysis also suggests that larger proportion of controls than treatment units leads to better balance for both of the methods.

Keywords: accuracy, covariate balances, generalized additive model, logistic regression, non-linearity, propensity score matching

Procedia PDF Downloads 343

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 157

Authors: Farid Khosravikia, Patricia Clayton

Abstract:

The main objective of this study is to evaluate the advantages and disadvantages of various machine learning techniques in forecasting ground-motion intensity measures given source characteristics, source-to-site distance, and local site condition. Intensity measures such as peak ground acceleration and velocity (PGA and PGV, respectively) as well as 5% damped elastic pseudospectral accelerations at different periods (PSA), are indicators of the strength of shaking at the ground surface. Estimating these variables for future earthquake events is a key step in seismic hazard assessment and potentially subsequent risk assessment of different types of structures. Typically, linear regression-based models, with pre-defined equations and coefficients, are used in ground motion prediction. However, due to the restrictions of the linear regression methods, such models may not capture more complex nonlinear behaviors that exist in the data. Thus, this study comparatively investigates potential benefits from employing other machine learning techniques as a statistical method in ground motion prediction such as Artificial Neural Network, Random Forest, and Support Vector Machine. The algorithms are adjusted to quantify event-to-event and site-to-site variability of the ground motions by implementing them as random effects in the proposed models to reduce the aleatory uncertainty. All the algorithms are trained using a selected database of 4,528 ground-motions, including 376 seismic events with magnitude 3 to 5.8, recorded over the hypocentral distance range of 4 to 500 km in Oklahoma, Kansas, and Texas since 2005. The main reason of the considered database stems from the recent increase in the seismicity rate of these states attributed to petroleum production and wastewater disposal activities, which necessities further investigation in the ground motion models developed for these states. Accuracy of the models in predicting intensity measures, generalization capability of the models for future data, as well as usability of the models are discussed in the evaluation process. The results indicate the algorithms satisfy some physically sound characteristics such as magnitude scaling distance dependency without requiring pre-defined equations or coefficients. Moreover, it is shown that, when sufficient data is available, all the alternative algorithms tend to provide more accurate estimates compared to the conventional linear regression-based method, and particularly, Random Forest outperforms the other algorithms. However, the conventional method is a better tool when limited data is available.

Keywords: artificial neural network, ground-motion models, machine learning, random forest, support vector machine

Procedia PDF Downloads 106

Authors: Alireza Vahid, Santu Rana, Sunil Gupta, Pratibha Vellanki, Svetha Venkatesh, Thomas Dorin

Abstract:

Aluminum is the preferred material for lightweight applications and its alloys are constantly improving. The high strength 7xxx alloys have been extensively used for structural components in aerospace and automobile industries for the past 50 years. In the next decade, a great number of airplanes will be retired, providing an obvious source of valuable used metals and great demand for cost-effective methods to re-use these alloys. The design of proper aerospace alloys is primarily based on optimizing strength and ductility, both of which can be improved by controlling the additional alloying elements as well as heat treatment conditions. In this project, we explore the design of high-performance alloys with 7xxx as a base material. These designed alloys have to be optimized and improved to compare with modern 7xxx-series alloys and to remain competitive for aircraft manufacturing. Aerospace alloys are extremely complex with multiple alloying elements and numerous processing steps making optimization often intensive and costly. In the present study, we used Bayesian optimization algorithm, a well-known adaptive design strategy, to optimize this multi-variable system. An Al alloy was proposed and the relevant heat treatment schedules were optimized, using the tensile yield strength as the output to maximize. The designed alloy has a maximum yield strength and ultimate tensile strength of more than 730 and 760 MPa, respectively, and is thus comparable to the modern high strength 7xxx-series alloys. The microstructure of this alloy is characterized by electron microscopy, indicating that the increased strength of the alloy is due to the presence of a high number density of refined precipitates.

Keywords: aluminum alloys, Bayesian optimization, heat treatment, tensile properties

Procedia PDF Downloads 100

Authors: Frank Kuebler, Rolf Steinhilper

Abstract:

Artificial neural networks (ANN) as well as Design of Experiments (DOE) based regression analysis (RA) are mainly used for modeling of complex systems. Both methodologies are commonly applied in process and quality control of manufacturing processes. Due to the fact that resource efficiency has become a critical concern for manufacturing companies, these models needs to be extended to predict resource-consumption of manufacturing processes. This paper describes an approach to use neural networks as well as DOE based regression analysis for predicting resource consumption of manufacturing processes and gives a comparison of the achievable results based on an industrial case study of a turning process.

Keywords: artificial neural network, design of experiments, regression analysis, resource efficiency, manufacturing process

Procedia PDF Downloads 504

Authors: Sajjad Baghernezhad

Abstract:

Meta-heuristic and hybrid algorithms have high adeer in modeling medical problems. In this study, a neural network was used to predict covid-19 among high-risk and low-risk patients. This study was conducted to collect the applied method and its target population consisting of 550 high-risk and low-risk patients from the Kerman University of medical sciences medical center to predict the coronavirus. In this study, the memetic algorithm, which is a combination of a genetic algorithm and a local search algorithm, has been used to update the weights of the neural network and develop the accuracy of the neural network. The initial study showed that the accuracy of the neural network was 88%. After updating the weights, the memetic algorithm increased by 93%. For the proposed model, sensitivity, specificity, positive predictivity value, value/accuracy to 97.4, 92.3, 95.8, 96.2, and 0.918, respectively; for the genetic algorithm model, 87.05, 9.20 7, 89.45, 97.30 and 0.967 and for logistic regression model were 87.40, 95.20, 93.79, 0.87 and 0.916. Based on the findings of this study, neural network models have a lower error rate in the diagnosis of patients based on individual variables and vital signs compared to the regression model. The findings of this study can help planners and health care providers in signing programs and early diagnosis of COVID-19 or Corona.

Keywords: COVID-19, decision support technique, neural network, genetic algorithm, memetic algorithm

Procedia PDF Downloads 54

Authors: M. Uysal, M. Yilmaz, I. Tiryakioğlu

Abstract:

Digital Terrain Model (DTM) is a digital numerical representation of the Earth's surface. DTMs have been applied to a diverse field of tasks, such as urban planning, military, glacier mapping, disaster management. In the expression of the Earth' surface as a mathematical model, an infinite number of point measurements are needed. Because of the impossibility of this case, the points at regular intervals are measured to characterize the Earth's surface and DTM of the Earth is generated. Hitherto, the classical measurement techniques and photogrammetry method have widespread use in the construction of DTM. At present, RADAR, LiDAR, and stereo satellite images are also used for the construction of DTM. In recent years, especially because of its superiorities, Airborne Light Detection and Ranging (LiDAR) has an increased use in DTM applications. A 3D point cloud is created with LiDAR technology by obtaining numerous point data. However recently, by the development in image mapping methods, the use of unmanned aerial vehicles (UAV) for photogrammetric data acquisition has increased DTM generation from image-based point cloud. The accuracy of the DTM depends on various factors such as data collection method, the distribution of elevation points, the point density, properties of the surface and interpolation methods. In this study, the random data reduction method is compared for DTMs generated from image based point cloud data. The original image based point cloud data set (100%) is reduced to a series of subsets by using random algorithm, representing the 75, 50, 25 and 5% of the original image based point cloud data set. Over the ANS campus of Afyon Kocatepe University as the test area, DTM constructed from the original image based point cloud data set is compared with DTMs interpolated from reduced data sets by Kriging interpolation method. The results show that the random data reduction method can be used to reduce the image based point cloud datasets to 50% density level while still maintaining the quality of DTM.

Keywords: DTM, Unmanned Aerial Vehicle (UAV), uniform, random, kriging

Procedia PDF Downloads 138

Authors: O. Delage, H. Bencherif, T. Portafaix, A. Bourdier

Abstract:

The main purpose of time series analysis is to learn about the dynamics behind some time ordered measurement data. Two approaches are used in the literature to get a better knowledge of the dynamics contained in observation data sequences. The first of these approaches concerns time series decomposition, which is an important analysis step allowing patterns and behaviors to be extracted as components providing insight into the mechanisms producing the time series. As in many cases, time series are short, noisy, and non-stationary. To provide components which are physically meaningful, methods such as Empirical Mode Decomposition (EMD), Empirical Wavelet Transform (EWT) or, more recently, Empirical Adaptive Wavelet Decomposition (EAWD) have been proposed. The second approach is to reconstruct the dynamics underlying the time series as a trajectory in state space by mapping a time series into a set of Rᵐ lag vectors by using the method of delays (MOD). Takens has proved that the trajectory obtained with the MOD technic is equivalent to the trajectory representing the dynamics behind the original time series. This work introduces the singular spectrum decomposition (SSD), which is a new adaptive method for decomposing non-linear and non-stationary time series in narrow-banded components. This method takes its origin from singular spectrum analysis (SSA), a nonparametric spectral estimation method used for the analysis and prediction of time series. As the first step of SSD is to constitute a trajectory matrix by embedding a one-dimensional time series into a set of lagged vectors, SSD can also be seen as a reconstruction method like MOD. We will first give a brief overview of the existing decomposition methods (EMD-EWT-EAWD). The SSD method will then be described in detail and applied to experimental time series of observations resulting from total columns of ozone measurements. The results obtained will be compared with those provided by the previously mentioned decomposition methods. We will also compare the reconstruction qualities of the observed dynamics obtained from the SSD and MOD methods.

Keywords: time series analysis, adaptive time series decomposition, wavelet, phase space reconstruction, singular spectrum analysis

Procedia PDF Downloads 85