Data Analysis and Applications 1

Christos H. Skiadas is the Founder and former Director of the Data Analysis and Forecasting Laboratory at the Technical University of Crete, Greece. He continues his work at the university at the ManLab in the Department of Production Engineering and Management.James R. Bozeman holds a PhD in Mathematics from Dartmouth College, USA, and is Professor of Mathematics at the American University of Malta.

• Preface xiIntroduction xvGilbert SAPORTAPart 1 Clustering and Regression 1Chapter 1 Cluster Validation by Measurement of Clustering Characteristics Relevant to the User 3Christian HENNIG1.1 Introduction 31.2 General notation 51.3 Aspects of cluster validity 61.3.1 Small within-cluster dissimilarities 61.3.2 Between-cluster separation 71.3.3 Representation of objects by centroids 71.3.4 Representation of dissimilarity structure by clustering 81.3.5 Small within-cluster gaps 91.3.6 Density modes and valleys 91.3.7 Uniform within-cluster density 121.3.8 Entropy 121.3.9 Parsimony 131.3.10 Similarity to homogeneous distributional shapes 131.3.11 Stability 131.3.12 Further Aspects 141.4 Aggregation of indexes 141.5 Random clusterings for calibrating indexes 151.5.1 Stupid K-centroids clustering 161.5.2 Stupid nearest neighbors clustering 161.5.3 Calibration 171.6 Examples 181.6.1 Artificial data set 181.6.2 Tetragonula bees data 201.7 Conclusion 221.8 Acknowledgment 231.9 References 23Chapter 2 Histogram-Based Clustering of Sensor Network Data 25Antonio BALZANELLA and Rosanna VERDE2.1 Introduction 252.2 Time series data stream clustering 282.2.1 Local clustering of histogram data 302.2.2 Online proximity matrix updating 322.2.3 Off-line partitioning through the dynamic clustering algorithm for dissimilarity tables 332.3 Results on real data 342.4 Conclusions 362.5 References 36Chapter 3 The Flexible Beta Regression Model 39Sonia MIGLIORATI, Agnese MDI BRISCO and Andrea ONGARO3.1 Introduction 393.2 The FB distribution 413.2.1 The beta distribution 413.2.2 The FB distribution 413.2.3 Reparameterization of the FB 423.3 The FB regression model 433.4 Bayesian inference 443.5 Illustrative application 473.6 Conclusion 483.7 References 50Chapter 4 S-weighted Instrumental Variables 53Jan Ámos VÍŠEK4.1 Summarizing the previous relevant results 534.2 The notations, framework, conditions and main tool 554.3 S-weighted estimator and its consistency 574.4 S-weighted instrumental variables and their consistency 594.5 Patterns of results of simulations 644.5.1 Generating the data 654.5.2 Reporting the results 664.6 Acknowledgment 694.7 References 69Part 2 Models and Modeling 73Chapter 5 Grouping Property and Decomposition of Explained Variance in Linear Regression 75Henri WALLARD5.1 Introduction 755.2 CAR scores 765.2.1 Definition and estimators 765.2.2 Historical criticism of the CAR scores 795.3 Variance decomposition methods and SVD 795.4 Grouping property of variance decomposition methods 805.4.1 Analysis of grouping property for CAR scores 815.4.2 Demonstration with two predictors 825.4.3 Analysis of grouping property using SVD 835.4.4 Application to the diabetes data set 865.5 Conclusions 875.6 References 88Chapter 6 On GARCH Models with Temporary Structural Changes 91Norio WATANABE and Fumiaki OKIHARA6.1 Introduction 916.2 The model 926.2.1 Trend model 926.2.2 Intervention GARCH model 936.3 Identification 966.4 Simulation 966.4.1 Simulation on trend model 966.4.2 Simulation on intervention trend model 986.5 Application 986.6 Concluding remarks 1026.7 References 103Chapter 7 A Note on the Linear Approximation of TAR Models 105Francesco GIORDANO, Marcella NIGLIO and Cosimo Damiano VITALE7.1 Introduction 1057.2 Linear representations and linear approximations of nonlinear models 1077.3 Linear approximation of the TAR model 1097.4 References 116Chapter 8 An Approximation of Social Well-Being Evaluation Using Structural Equation Modeling 117Leonel SANTOS-BARRIOS, Monica RUIZ-TORRES, William GÓMEZ-DEMETRIO, Ernesto SÁNCHEZ-VERA, Ana LORGA DA SILVA and Francisco MARTÍNEZ-CASTAÑEDA8.1 Introduction 1178.2 Wellness1188.3 Social welfare 1188.4 Methodology 1198.5 Results 1208.6 Discussion 1238.7 Conclusions 1238.8 References 123Chapter 9 An SEM Approach to Modeling Housing Values 125Jim FREEMAN and Xin ZHAO9.1 Introduction 1259.2 Data 1269.3 Analysis 1279.4 Conclusions 1349.5 References 135Chapter 10 Evaluation of Stopping Criteria for Ranks in Solving Linear Systems 137Benard ABOLA, Pitos BIGANDA, Christopher ENGSTRÖM and Sergei SILVESTROV10.1 Introduction 13710.2 Methods 13910.2.1 Preliminaries 13910.2.2 Iterative methods 14010.3 Formulation of linear systems 14210.4 Stopping criteria 14310.5 Numerical experimentation of stopping criteria 14610.5.1 Convergence of stopping criterion 14710.5.2 Quantiles 14710.5.3 Kendall correlation coefficient as stopping criterion 14810.6 Conclusions 15010.7 Acknowledgments 15110.8 References 151Chapter 11 Estimation of a Two-Variable Second-Degree Polynomial via Sampling 153Ioanna PAPATSOUMA, Nikolaos FARMAKIS and Eleni KETZAKI11.1 Introduction 15311.2 Proposed method 15411.2.1 First restriction 15411.2.2 Second restriction 15511.2.3 Third restriction 15611.2.4 Fourth restriction 15611.2.5 Fifth restriction 15711.2.6 Coefficient estimates 15811.3 Experimental approaches 15911.3.1 Experiment A 15911.3.2 Experiment B 16111.4 Conclusions 16311.5 References 163Part 3 Estimators, Forecasting and Data Mining 165Chapter 12 Displaying Empirical Distributions of Conditional Quantile Estimates: An Application of Symbolic Data Analysis to the Cost Allocation Problem in Agriculture 167Dominique DESBOIS12.1 Conceptual framework and methodological aspects of cost allocation 16712.2 The empirical model of specific production cost estimates 16812.3 The conditional quantile estimation 16912.4 Symbolic analyses of the empirical distributions of specific costs 17012.5 The visualization and the analysis of econometric results 17212.6 Conclusion 17812.7 Acknowledgments 17912.8 References 179Chapter 13 Frost Prediction in Apple Orchards Based upon Time Series Models 181Monika ATOMKOWICZ and Armin OSCHMITT13.1 Introduction 18113.2 Weather database 18213.3 ARIMA forecast model 18313.3.1 Stationarity and differencing 18413.3.2 Non-seasonal ARIMA models 18613.4 Model building 18813.4.1 ARIMA and LR models 18813.4.2 Binary classification of the frost data 18913.4.3 Training and test set 18913.5 Evaluation 18913.6 ARIMA model selection 19013.7 Conclusions 19213.8 Acknowledgments 19313.9 References 193Chapter 14 Efficiency Evaluation of Multiple-Choice Questions and Exams 195Evgeny GERSHIKOV and Samuel KOSOLAPOV14.1 Introduction 19514.2 Exam efficiency evaluation 19614.2.1 Efficiency measures and efficiency weighted grades 19614.2.2 Iterative execution 19814.2.3 Postprocessing 19914.3 Real-life experiments and results 20014.4 Conclusions 20314.5 References 204Chapter 15 Methods of Modeling and Estimation in Mortality 205Christos HSKIADAS and Konstantinos NZAFEIRIS15.1 Introduction 20515.2 The appearance of life tables 20615.3 On the law of mortality 20715.4 Mortality and health 21115.5 An advanced health state function form 21715.6 Epilogue 22015.7 References 221Chapter 16 An Application of Data Mining Methods to the Analysis of Bank Customer Profitability and Buying Behavior 225Pedro GODINHO, Joana DIAS and Pedro TORRES16.1 Introduction 22516.2 Data set 22716.3 Short-term forecasting of customer profitability 23016.4 Churn prediction 23516.5 Next-product-to-buy 23616.6 Conclusions and future research 23816.7 References 239List of Authors 241Index 245