Search results for: copula-based regression models

Authors: Mukesh Khadka, Alexander Paz, Hanns de la Fuente-Mella

Abstract:

Typically, pavement performance models are developed in two steps: (i) pavement segments with similar characteristics are grouped together to form a cluster, and (ii) the corresponding performance models are developed using statistical techniques. A challenge is to select the characteristics that define clusters and the segments associated with them. If inappropriate characteristics are used, clusters may include homogeneous segments with different performance behavior or heterogeneous segments with similar performance behavior. Prediction accuracy of performance models can be improved by grouping the pavement segments into more uniform clusters by including both characteristics and a performance measure. This grouping is not always possible due to limited information. It is impractical to include all the potential significant factors because some of them are potentially unobserved or difficult to measure. Historical performance of pavement segments could be used as a proxy to incorporate the effect of the missing potential significant factors in clustering process. The current state-of-the-art proposes Clusterwise Linear Regression (CLR) to determine the pavement clusters and the associated performance models simultaneously. CLR incorporates the effect of significant factors as well as a performance measure. In this study, a mathematical program was formulated for CLR models including multiple explanatory variables. Pavement data collected recently over the entire state of Nevada were used. International Roughness Index (IRI) was used as a pavement performance measure because it serves as a unified standard that is widely accepted for evaluating pavement performance, especially in terms of riding quality. Results illustrate the advantage of the using CLR. Previous studies have used CLR along with experimental data. This study uses actual field data collected across a variety of environmental, traffic, design, and construction and maintenance conditions.

Keywords: clusterwise regression, pavement management system, performance model, optimization

Procedia PDF Downloads 227

Authors: Masood Beheshtirad

Abstract:

Identification of regions having potential for landslide occurrence is one of the basic measures in natural resources management. Different landslide hazard mapping models are proposed based on the environmental condition and goals. In this research landslide hazard map using multiple regression model were provided and applicability of this model is investigated in Baghdasht watershed. Dependent variable is landslide inventory map and independent variables consist of information layers as Geology, slope, aspect, distance from river, distance from road, fault and land use. For doing this, existing landslides have been identified and an inventory map made. The landslide hazard map is based on the multiple regression provided. The level of similarity potential hazard classes and figures of this model were compared with the landslide inventory map in the SPSS environments. Results of research showed that there is a significant correlation between the potential hazard classes and figures with area of the landslides. The multiple regression model is suitable for application in the Baghdasht Watershed.

Keywords: landslide, mapping, multiple model, regression

Procedia PDF Downloads 302

Authors: Deyadeen Ali Alshibani

Abstract:

In the literature, the multi-armed bandit problem as a statistical decision model of an agent trying to optimize his decisions while improving his information at the same time. There are several different algorithms models and their applications on this problem. In this paper, we evaluate the Regret-regression through comparing with Q-learning method. A simulation on determination of optimal treatment regime is presented in detail.

Keywords: optimal, bandit problem, optimization, dynamic programming

Procedia PDF Downloads 424

Authors: Krishang Surapaneni

Abstract:

The goal of this project is to determine how to predict important aspects of options, including the ask price. We want to compare different machine learning models to learn the best model and the best hyperparameters for that model for this purpose and data set. Option pricing is a relatively new field, and it can be very complicated and intimidating, especially to inexperienced people, so we want to create a machine learning model that can predict important aspects of an option stock, which can aid in future research. We tested multiple different models and experimented with hyperparameter tuning, trying to find some of the best parameters for a machine-learning model. We tested three different models: a Random Forest Regressor, a linear regressor, and an MLP (multi-layer perceptron) regressor. The most important feature in this experiment is the ask price; this is what we were trying to predict. In the field of stock pricing prediction, there is a large potential for error, so we are unable to determine the accuracy of the models based on if they predict the pricing perfectly. Due to this factor, we determined the accuracy of the model by finding the average percentage difference between the predicted and actual values. We tested the accuracy of the machine learning models by comparing the actual results in the testing data and the predictions made by the models. The linear regression model performed worst, with an average percentage error of 17.46%. The MLP regressor had an average percentage error of 11.45%, and the random forest regressor had an average percentage error of 7.42%

Keywords: finance, linear regression model, machine learning model, neural network, stock price

Procedia PDF Downloads 54

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 493

Authors: Anita Setianingrum, Oki S. Jaya, Zuherman Rustam

Abstract:

Forecasting stock price is a challenging task due to the complex time series of the data. The complexity arises from many variables that affect the stock market. Many time series models have been proposed before, but those previous models still have some problems: 1) put the subjectivity of choosing the technical indicators, and 2) rely upon some assumptions about the variables, so it is limited to be applied to all datasets. Therefore, this paper studied a novel Adaptive Neuro-Fuzzy Inference System (ANFIS) time series model based on Support Vector Regression (SVR) for forecasting the stock market. In order to evaluate the performance of proposed models, stock market transaction data of TAIEX and HIS from January to December 2015 is collected as experimental datasets. As a result, the method has outperformed its counterparts in terms of accuracy.

Keywords: ANFIS, fuzzy time series, stock forecasting, SVR

Procedia PDF Downloads 217

Authors: Adriano Z. Zambom, Preethi Ravikumar

Abstract:

One of the biggest challenges in nonparametric regression is the curse of dimensionality. Additive models are known to overcome this problem by estimating only the individual additive effects of each covariate. However, if the model is misspecified, the accuracy of the estimator compared to the fully nonparametric one is unknown. In this work the efficiency of completely nonparametric regression estimators such as the Loess is compared to the estimators that assume additivity in several situations, including additive and non-additive regression scenarios. The comparison is done by computing the oracle mean square error of the estimators with regards to the true nonparametric regression function. Then, a backward elimination selection procedure based on the Akaike Information Criteria is proposed, which is computed from either the additive or the nonparametric model. Simulations show that if the additive model is misspecified, the percentage of time it fails to select important variables can be higher than that of the fully nonparametric approach. A dimension reduction step is included when nonparametric estimator cannot be computed due to the curse of dimensionality. Finally, the Boston housing dataset is analyzed using the proposed backward elimination procedure and the selected variables are identified.

Keywords: additive model, nonparametric regression, variable selection, Akaike Information Criteria

Procedia PDF Downloads 241

Authors: Florin Leon, Silvia Curteanu

Abstract:

Developing complete mechanistic models for polymerization reactors is not easy, because complex reactions occur simultaneously; there is a large number of kinetic parameters involved and sometimes the chemical and physical phenomena for mixtures involving polymers are poorly understood. To overcome these difficulties, empirical models based on sampled data can be used instead, namely regression methods typical of machine learning field. They have the ability to learn the trends of a process without any knowledge about its particular physical and chemical laws. Therefore, they are useful for modeling complex processes, such as the free radical polymerization of methyl methacrylate achieved in a batch bulk process. The goal is to generate accurate predictions of monomer conversion, numerical average molecular weight and gravimetrical average molecular weight. This process is associated with non-linear gel and glass effects. For this purpose, an adaptive sampling technique is presented, which can select more samples around the regions where the values have a higher variation. Several machine learning methods are used for the modeling and their performance is compared: support vector machines, k-nearest neighbor, k-nearest neighbor and random forest, as well as an original algorithm, large margin nearest neighbor regression. The suggested method provides very good results compared to the other well-known regression algorithms.

Keywords: batch bulk methyl methacrylate polymerization, adaptive sampling, machine learning, large margin nearest neighbor regression

Procedia PDF Downloads 270

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

Abstract:

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

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

Procedia PDF Downloads 58

Authors: Milica Ž. Karadžić, Lidija R. Jevrić, Sanja Podunavac-Kuzmanović, Strahinja Z. Kovačević, Anamarija I. Mandić, Katarina Penov-Gaši, Andrea R. Nikolić, Aleksandar M. Oklješa

Abstract:

This study considered selection of the in silico molecular descriptors and the models for newly synthesized steroid derivatives description and their characterization using chemometric techniques. Multiple linear regression (MLR) models were established and gave the best molecular descriptors for quantitative structure-retention relationship (QSRR) modeling of the retention of the investigated molecules. MLR models were without multicollinearity among the selected molecular descriptors according to the variance inflation factor (VIF) values. Used molecular descriptors were ranked using generalized pair correlation method (GPCM). In this method, the significant difference between independent variables can be noticed regardless almost equal correlation between dependent variable. Generated MLR models were statistically and cross-validated and the best models were kept. Models were ranked using sum of ranking differences (SRD) method. According to this method, the most consistent QSRR model can be found and similarity or dissimilarity between the models could be noticed. In this study, SRD was performed using average values of experimentally observed data as a golden standard. Chemometric analysis was conducted in order to characterize newly synthesized steroid derivatives for further investigation regarding their potential biological activity and further synthesis. This article is based upon work from COST Action (CM1105), supported by COST (European Cooperation in Science and Technology).

Keywords: generalized pair correlation method, molecular descriptors, regression analysis, steroids, sum of ranking differences

Procedia PDF Downloads 313

Authors: Animish Jain

Abstract:

This study deals with using Multiplicative Weight Update within artificial intelligence and machine learning to create models that can diagnose skin cancer using microscopic images of cancer samples. In this study, the multiplicative weight update method is used to take the predictions of multiple models to try and acquire more accurate results. Logistic Regression, Convolutional Neural Network (CNN), and Support Vector Machine Classifier (SVMC) models are employed within the Multiplicative Weight Update system. These models are trained on pictures of skin cancer from the ISIC-Archive, to look for patterns to label unseen scans as either benign or malignant. These models are utilized in a multiplicative weight update algorithm which takes into account the precision and accuracy of each model through each successive guess to apply weights to their guess. These guesses and weights are then analyzed together to try and obtain the correct predictions. The research hypothesis for this study stated that there would be a significant difference in the accuracy of the three models and the Multiplicative Weight Update system. The SVMC model had an accuracy of 77.88%. The CNN model had an accuracy of 85.30%. The Logistic Regression model had an accuracy of 79.09%. Using Multiplicative Weight Update, the algorithm received an accuracy of 72.27%. The final conclusion that was drawn was that there was a significant difference in the accuracy of the three models and the Multiplicative Weight Update system. The conclusion was made that using a CNN model would be the best option for this problem rather than a Multiplicative Weight Update system. This is due to the possibility that Multiplicative Weight Update is not effective in a binary setting where there are only two possible classifications. In a categorical setting with multiple classes and groupings, a Multiplicative Weight Update system might become more proficient as it takes into account the strengths of multiple different models to classify images into multiple categories rather than only two categories, as shown in this study. This experimentation and computer science project can help to create better algorithms and models for the future of artificial intelligence in the medical imaging field.

Keywords: artificial intelligence, machine learning, multiplicative weight update, skin cancer

Procedia PDF Downloads 45

Authors: Prashant Verma, Prafulla K. Swain, K. K. Singh, Mukti Khetan

Abstract:

Objective: In India, around 20,000 women die every year due to abortion-related complications. In the modelling of count variables, there is sometimes a preponderance of zero counts. This article concerns the estimation of various count regression models to predict the average number of spontaneous abortion among women in the Punjab state of India. It also assesses the factors associated with the number of spontaneous abortions. Materials and methods: The study included 27,173 married women of Punjab obtained from the DLHS-4 survey (2012-13). Poisson regression (PR), Negative binomial (NB) regression, zero hurdle negative binomial (ZHNB), and zero-inflated negative binomial (ZINB) models were employed to predict the average number of spontaneous abortions and to identify the determinants affecting the number of spontaneous abortions. Results: Statistical comparisons among four estimation methods revealed that the ZINB model provides the best prediction for the number of spontaneous abortions. Antenatal care (ANC) place, place of residence, total children born to a woman, woman's education and economic status were found to be the most significant factors affecting the occurrence of spontaneous abortion. Conclusions: The study offers a practical demonstration of techniques designed to handle count variables. Statistical comparisons among four estimation models revealed that the ZINB model provided the best prediction for the number of spontaneous abortions and is recommended to be used to predict the number of spontaneous abortions. The study suggests that women receive institutional Antenatal care to attain limited parity. It also advocates promoting higher education among women in Punjab, India.

Keywords: count data, spontaneous abortion, Poisson model, negative binomial model, zero hurdle negative binomial, zero-inflated negative binomial, regression

Procedia PDF Downloads 125

Authors: Deepika Christopher, Garima Anand

Abstract:

To determine the best model for churn prediction in the telecom industry, this paper compares 11 machine learning algorithms, namely Logistic Regression, Support Vector Machine, Random Forest, Decision Tree, XGBoost, LightGBM, Cat Boost, AdaBoost, Extra Trees, Deep Neural Network, and Hybrid Model (MLPClassifier). It also aims to pinpoint the top three factors that lead to customer churn and conducts customer segmentation to identify vulnerable groups. According to the data, the Logistic Regression model performs the best, with an F1 score of 0.6215, 81.76% accuracy, 68.95% precision, and 56.57% recall. The top three attributes that cause churn are found to be tenure, Internet Service Fiber optic, and Internet Service DSL; conversely, the top three models in this article that perform the best are Logistic Regression, Deep Neural Network, and AdaBoost. The K means algorithm is applied to establish and analyze four different customer clusters. This study has effectively identified customers that are at risk of churn and may be utilized to develop and execute strategies that lower customer attrition.

Keywords: attrition, retention, predictive modeling, customer segmentation, telecommunications

Procedia PDF Downloads 25

Authors: Atheer Al-Nuaimi, Harry Evdorides

Abstract:

Every day many people die or get disabled or injured on roads around the world, which necessitates more specific treatments for transportation safety issues. International road assessment program (iRAP) model is one of the comprehensive road safety models which accounting for many factors that affect road safety in a cost-effective way in low and middle income countries. In iRAP model road safety has been divided into five star ratings from 1 star (the lowest level) to 5 star (the highest level). These star ratings are based on star rating score which is calculated by iRAP methodology depending on road attributes, traffic volumes and operating speeds. The outcome of iRAP methodology are the treatments that can be used to improve road safety and reduce fatalities and serious injuries (FSI) numbers. These countermeasures can be used separately as a single countermeasure or mix as multiple countermeasures for a location. There is general agreement that the adequacy of a countermeasure is liable to consistent losses when it is utilized as a part of mix with different countermeasures. That is, accident diminishment appraisals of individual countermeasures cannot be easily added together. The iRAP model philosophy makes utilization of a multiple countermeasure adjustment factors to predict diminishments in the effectiveness of road safety countermeasures when more than one countermeasure is chosen. A multiple countermeasure correction factors are figured for every 100-meter segment and for every accident type. However, restrictions of this methodology incorporate a presumable over-estimation in the predicted crash reduction. This study aims to adjust this correction factor by developing new models to calculate the effect of using multiple countermeasures on the number of fatalities for a location or an entire road. Regression models have been used to establish relationships between crash frequencies and the factors that affect their rates. Multiple linear regression, negative binomial regression, and Poisson regression techniques were used to develop models that can address the effectiveness of using multiple countermeasures. Analyses are conducted using The R Project for Statistical Computing showed that a model developed by negative binomial regression technique could give more reliable results of the predicted number of fatalities after the implementation of road safety multiple countermeasures than the results from iRAP model. The results also showed that the negative binomial regression approach gives more precise results in comparison with multiple linear and Poisson regression techniques because of the overdispersion and standard error issues.

Keywords: international road assessment program, negative binomial, road multiple countermeasures, road safety

Procedia PDF Downloads 212

Authors: Bader Alkandari, Jamal Y. Madouh, Ahmad M. Alkandari, Anwar A. Alnaqi

Abstract:

A new formulation of fuzzy linear estimation problem is presented. It is formulated as a linear programming problem. The objective is to minimize the spread of the data points, taking into consideration the type of the membership function of the fuzzy parameters to satisfy the constraints on each measurement point and to insure that the original membership is included in the estimated membership. Different models are developed for a fuzzy triangular membership. The proposed models are applied to different examples from the area of fuzzy linear regression and finally to different examples for estimating the electrical load on a busbar. It had been found that the proposed technique is more suited for electrical load estimation, since the nature of the load is characterized by the uncertainty and vagueness.

Keywords: fuzzy regression, load estimation, fuzzy linear parameters, electrical load estimation

Procedia PDF Downloads 510

Authors: Tahani Salman Alangari

Abstract:

Student success can be impacted by internal factors such as their emotional well-being and external factors such as organizational support and instructional support in the classroom. This study is to identify at least one factor that forecasts student engagement. It is a cross-sectional, conducted on 6206 teachers and encompassed three years of data collection and observations of math instruction in approximately 50 schools and 300 classrooms. A multiple linear regression revealed that a model predicting student engagement from emotional support, classroom organization, and instructional support was significant. Four linear regression models were tested using hierarchical regression to examine the effects of independent variables: emotional support was the highest predictor of student engagement while instructional support was the lowest.

Keywords: student engagement, emotional support, organizational support, instructional support, well-being

Procedia PDF Downloads 51

Authors: Mohammad Anwar, Shah Waliullah

Abstract:

This study investigated panel data regression models. This paper used Bayesian and classical methods to study the impact of institutions on economic growth from data (1990-2014), especially in developing countries. Under the classical and Bayesian methodology, the two-panel data models were estimated, which are common effects and fixed effects. For the Bayesian approach, the prior information is used in this paper, and normal gamma prior is used for the panel data models. The analysis was done through WinBUGS14 software. The estimated results of the study showed that panel data models are valid models in Bayesian methodology. In the Bayesian approach, the effects of all independent variables were positively and significantly affected by the dependent variables. Based on the standard errors of all models, we must say that the fixed effect model is the best model in the Bayesian estimation of panel data models. Also, it was proved that the fixed effect model has the lowest value of standard error, as compared to other models.

Keywords: Bayesian approach, common effect, fixed effect, random effect, Dynamic Random Effect Model

Procedia PDF Downloads 52

Authors: Endang Jati Mat Sahid, Hussain Ali Bekhet

Abstract:

Ability to supply reliable and secure electricity has been one of the crucial components of economic development for any country. Forecasting of electricity production is therefore very important for accurate investment planning of generation power plants. In this study, we aim to examine and analyze the factors that affect electricity generation. Multiple regression models were used to find the relationship between various variables and electricity production. The models will simultaneously determine the effects of the variables on electricity generation. Many variables influencing electricity generation, i.e. natural gas (NG), coal (CO), fuel oil (FO), renewable energy (RE), gross domestic product (GDP) and fuel prices (FP), were examined for Malaysia. The results demonstrate that NG, CO, and FO were the main factors influencing electricity generation growth. This study then identified a number of policy implications resulting from the empirical results.

Keywords: energy policy, energy security, electricity production, Malaysia, the regression model

Procedia PDF Downloads 126

Authors: K. K. Lee, H. W. Han, H. L. Kang, T. A. Kim, S. H. Han

Abstract:

For the robust optimization of the manufacturing product design, there are design objectives that must be achieved, such as a minimization of the mean and standard deviation in objective functions within the required sensitivity constraints. The authors utilized the sensitivity of objective functions and constraints with respect to the effective design variables to reduce the computational burden associated with the evaluation of the probabilities. The individual mean and sensitivity values could be estimated easily by using the 9 level orthogonal array based response surface models optimized by the stepwise regression. The present study evaluates a proposed procedure from the robust optimization of rubber domes that are commonly used for keyboard switching, by using the 9 level orthogonal array and stepwise regression along with a desirability function. In addition, a new robust optimization process, i.e., the I2GEO (Identify, Integrate, Generate, Explore and Optimize), was proposed on the basis of the robust optimization in rubber domes. The optimized results from the response surface models and the estimated results by using the finite element analysis were consistent within a small margin of error. The standard deviation of objective function is decreasing 54.17% with suggested sensitivity based robust optimization. (Business for Cooperative R&D between Industry, Academy, and Research Institute funded Korea Small and Medium Business Administration in 2017, S2455569)

Keywords: objective function, orthogonal array, response surface model, robust optimization, stepwise regression

Procedia PDF Downloads 261

Authors: Daniel Fundi Murithi

Abstract:

Missing data is a real bane in many surveys. To overcome the problems caused by missing data, partial deletion, and single imputation methods, among others, have been proposed. However, problems such as discarding usable data and inaccuracy in reproducing known population parameters and standard errors are associated with them. For regression and stochastic imputation, it is assumed that there is a variable with complete cases to be used as a predictor in estimating missing values in the other variable, and the relationship between the two variables is linear, which might not be realistic in practice. In this project, we estimate population total in presence of missing values in two-phase sampling. Instead of regression or stochastic models, non-parametric model based regression model is used in imputing missing values. Empirical study showed that nonparametric model-based regression imputation is better in reproducing variance of population total estimate obtained when there were no missing values compared to mean, median, regression, and stochastic imputation methods. Although regression and stochastic imputation were better than nonparametric model-based imputation in reproducing population total estimates obtained when there were no missing values in one of the sample sizes considered, nonparametric model-based imputation may be used when the relationship between outcome and predictor variables is not linear.

Keywords: finite population total, missing data, model-based imputation, two-phase sampling

Procedia PDF Downloads 104

Authors: Sifriyani Sifriyani, I Nyoman Budiantara, Sri Haryatmi, Gunardi Gunardi

Abstract:

East Java Province has a first rank as a province that has the most counties and cities in Indonesia and has the largest population. In 2015, the population reached 38.847.561 million, this figure showed a very high population growth. High population growth is feared to lead to increase the levels of unemployment. In this study, the researchers mapped and modeled the unemployment rate with 6 variables that were supposed to influence. Modeling was done by nonparametric geographically weighted regression methods with truncated spline approach. This method was chosen because spline method is a flexible method, these models tend to look for its own estimation. In this modeling, there were point knots, the point that showed the changes of data. The selection of the optimum point knots was done by selecting the most minimun value of Generalized Cross Validation (GCV). Based on the research, 6 variables were declared to affect the level of unemployment in eastern Java. They were the percentage of population that is educated above high school, the rate of economic growth, the population density, the investment ratio of total labor force, the regional minimum wage and the ratio of the number of big industry and medium scale industry from the work force. The nonparametric geographically weighted regression models with truncated spline approach had a coefficient of determination 98.95% and the value of MSE equal to 0.0047.

Keywords: East Java, nonparametric geographically weighted regression, spatial, spline approach, unemployed rate

Procedia PDF Downloads 290

Authors: Jay L. Fu

Abstract:

Patients with Alzheimer's disease progressively lose their memory and thinking skills and, eventually, the ability to carry out simple daily tasks. The disease is irreversible, but early detection and treatment can slow down the disease progression. In this research, publicly available MRI data and demographic data from 373 MRI imaging sessions were utilized to build models to predict dementia. Various machine learning models, including logistic regression, k-nearest neighbor, support vector machine, random forest, and neural network, were developed. Data were divided into training and testing sets, where training sets were used to build the predictive model, and testing sets were used to assess the accuracy of prediction. Key risk factors were identified, and various models were compared to come forward with the best prediction model. Among these models, the random forest model appeared to be the best model with an accuracy of 90.34%. MMSE, nWBV, and gender were the three most important contributing factors to the detection of Alzheimer’s. Among all the models used, the percent in which at least 4 of the 5 models shared the same diagnosis for a testing input was 90.42%. These machine learning models allow early detection of Alzheimer’s with good accuracy, which ultimately leads to early treatment of these patients.

Keywords: Alzheimer's disease, clinical diagnosis, magnetic resonance imaging, machine learning prediction

Procedia PDF Downloads 117

Authors: Katawut Kaewbanjong

Abstract:

We analyzed a volume of data and found significant user satisfaction in software project factors. A statistical significance analysis (logistic regression) and collinearity analysis determined the significance factors from a group of 71 pre-defined factors from 191 software projects in ISBSG Release 12. The eight prediction models used for testing the prediction potential of these factors were Neural network, k-NN, Naïve Bayes, Random forest, Decision tree, Gradient boosted tree, linear regression and logistic regression prediction model. Fifteen pre-defined factors were truly significant in predicting user satisfaction, and they provided 82.71% prediction accuracy when used with a neural network prediction model. These factors were client-server, personnel changes, total defects delivered, project inactive time, industry sector, application type, development type, how methodology was acquired, development techniques, decision making process, intended market, size estimate approach, size estimate method, cost recording method, and effort estimate method. These findings may benefit software development managers considerably.

Keywords: prediction model, statistical analysis, software project, user satisfaction factor

Procedia PDF Downloads 90

Authors: Y. H. Wu, Ji-Yuan Lin, Yu-Ming Liou

Abstract:

This study aims to set up the landslide susceptibility map of County Route 89 at Ren-Ai Township in Nantou County using the Instability Index Method and Logistic regression. Seven susceptibility factors including Slope Angle, Aspect, Elevation, Distance to fold, Distance to River, Distance to Road and Accumulated Rainfall were obtained by GIS based on the Typhoon Toraji landslide area identified by Industrial Technology Research Institute in 2001. To calculate the landslide percentage of each factor and acquire the weight and grade the grid by means of Instability Index Method. In this study, landslide susceptibility can be classified into four grades: high, medium high, medium low and low, in order to determine the advantages and disadvantages of the two models. The precision of this model is verified by classification error matrix and SRC curve. These results suggest that the logistic regression model is a preferred method than instability index in the assessment of landslide susceptibility. It is suitable for the landslide prediction and precaution in this area in the future.

Keywords: instability index method, logistic regression, landslide susceptibility, SRC curve

Procedia PDF Downloads 259

Authors: Darren Zou

Abstract:

Forecasting the stock market is a very challenging task. Various economic indicators such as GDP, exchange rates, interest rates, and unemployment have a substantial impact on the stock market. Time series models are the traditional methods used to predict stock market changes. In this paper, a machine learning method, Bayesian Additive Regression Tree (BART) is used in predicting stock market indexes based on multiple economic indicators. BART can be used to model heterogeneous treatment effects, and thereby works well when models are misspecified. It also has the capability to handle non-linear main effects and multi-way interactions without much input from financial analysts. In this research, BART is proposed to provide a reliable prediction on day-to-day stock market activities. By comparing the analysis results from BART and with time series method, BART can perform well and has better prediction capability than the traditional methods.

Keywords: BART, Bayesian, predict, stock

Procedia PDF Downloads 97

Authors: Petr Gurný

Abstract:

One of the most important tasks in the risk management is the correct determination of probability of default (PD) of particular financial subjects. In this paper a possibility of determination of financial institution’s PD according to the credit-scoring models is discussed. The paper is divided into the two parts. The first part is devoted to the estimation of the three different models (based on the linear discriminant analysis, logit regression and probit regression) from the sample of almost three hundred US commercial banks. Afterwards these models are compared and verified on the control sample with the view to choose the best one. The second part of the paper is aimed at the application of the chosen model on the portfolio of three key Czech banks to estimate their present financial stability. However, it is not less important to be able to estimate the evolution of PD in the future. For this reason, the second task in this paper is to estimate the probability distribution of the future PD for the Czech banks. So, there are sampled randomly the values of particular indicators and estimated the PDs’ distribution, while it’s assumed that the indicators are distributed according to the multidimensional subordinated Lévy model (Variance Gamma model and Normal Inverse Gaussian model, particularly). Although the obtained results show that all banks are relatively healthy, there is still high chance that “a financial crisis” will occur, at least in terms of probability. This is indicated by estimation of the various quantiles in the estimated distributions. Finally, it should be noted that the applicability of the estimated model (with respect to the used data) is limited to the recessionary phase of the financial market.

Keywords: credit-scoring models, multidimensional subordinated Lévy model, probability of default

Procedia PDF Downloads 424

Authors: Yanqi Yang, Chongshan Liao, Cheuk Hin Ho, Susan Bridges

Abstract:

Aim: To investigate student’s perceptions of using e-models in an inquiry-based curriculum. Approach: 52 second-year dental students completed a pre- and post-test questionnaire relating to their perceptions of e-models and their use in inquiry-based learning. The pre-test occurred prior to any learning with e-models. The follow-up survey was conducted after one year's experience of using e-models. Results: There was no significant difference between the two sets of questionnaires regarding student’s perceptions of the usefulness of e-models and their willingness to use e-models in future inquiry-based learning. Most of the students preferred using both plaster models and e-models in tandem. Conclusion: Students did not change their attitude towards e-models and most of them agreed or were neutral that e-models are useful in inquiry-based learning. Whilst recognizing the utility of 3D models for learning, student's preference for combining these with solid models has implications for the development of haptic sensibility in an operative discipline.

Keywords: e-models, inquiry-based curriculum, education, questionnaire

Procedia PDF Downloads 397

Authors: Marius Gavrilescu, Sabina-Adriana Floria, Florin Leon, Silvia Curteanu, Costel Anton

Abstract:

Natural sciences provide a wide range of experimental data whose related problems require study and modeling beyond the capabilities of conventional methodologies. Such problems have solution spaces whose complexity and high dimensionality require correspondingly complex regression methods for proper characterization. In this context, we propose an optimization method which consists in a hybrid dual optimizer setup: a global optimizer based on a modified variant of the popular Imperialist Competitive Algorithm (ICA), and a local optimizer based on a gradient descent approach. The ICA is modified such that intermediate solution populations are more quickly and efficiently pruned of low-fitness individuals by appropriately altering the assimilation, revolution and competition phases, which, combined with an initialization strategy based on low-discrepancy sampling, allows for a more effective exploration of the corresponding solution space. Subsequently, gradient-based optimization is used locally to seek the optimal solution in the neighborhoods of the solutions found through the modified ICA. We use this combined approach to find the optimal configuration and weights of a fully-connected neural network, resulting in regression models used to characterize the process of obtained bricks using silicon-based materials. Installations in the raw ceramics industry, i.e., bricks, are characterized by significant energy consumption and large quantities of emissions. Thus, the purpose of our approach is to determine by simulation the working conditions, including the manufacturing mix recipe with the addition of different materials, to minimize the emissions represented by CO and CH4. Our approach determines regression models which perform significantly better than those found using the traditional ICA for the aforementioned problem, resulting in better convergence and a substantially lower error.

Keywords: optimization, biologically inspired algorithm, regression models, bricks, emissions

Procedia PDF Downloads 55

Authors: Gianna Zou

Abstract:

Propensity score matching (PSM), introduced by Paul R. Rosenbaum and Donald Rubin in 1983, is a popular statistical matching technique which tries to estimate the treatment effects by taking into account covariates that could impact the efficacy of study medication in clinical trials. PSM can be used to reduce the bias due to confounding variables. However, PSM assumes that the response values are normally distributed. In some cases, this assumption may not be held. In this paper, a machine learning method - Bayesian Additive Regression Tree (BART), is used as a more robust method of matching. BART can work well when models are misspecified since it can be used to model heterogeneous treatment effects. Moreover, it has the capability to handle non-linear main effects and multiway interactions. In this research, a BART Matching Method (BMM) is proposed to provide a more reliable matching method over PSM. By comparing the analysis results from PSM and BMM, BMM can perform well and has better prediction capability when the response values are not normally distributed.

Keywords: BART, Bayesian, matching, regression

Procedia PDF Downloads 115

Authors: Shiwei Deng, Yang Bao

Abstract:

In this paper, we present a model-based regression test suite reducing approach that uses EFSM model dependence analysis and probability-driven greedy algorithm to reduce software regression test suites. The approach automatically identifies the difference between the original model and the modified model as a set of elementary model modifications. The EFSM dependence analysis is performed for each elementary modification to reduce the regression test suite, and then the probability-driven greedy algorithm is adopted to select the minimum set of test cases from the reduced regression test suite that cover all interaction patterns. Our initial experience shows that the approach may significantly reduce the size of regression test suites.

Keywords: dependence analysis, EFSM model, greedy algorithm, regression test

Procedia PDF Downloads 399