A User's Guide to Business Analytics

A User's Guide to Business Analytics provides a comprehensive discussion of statistical methods useful to the business analyst. Methods are developed from a fairly basic level to accommodate readers who have limited training in the theory of statistics. A substantial number of case studies and numerical illustrations using the R-software package are provided for the benefit of motivated beginners who want to get a head start in analytics as well as for experts on the job who will benefit by using this text as a reference book. The book is comprised of 12 chapters. The first chapter focuses on business analytics, along with its emergence and application, and sets up a context for the whole book. The next three chapters introduce R and provide a comprehensive discussion on descriptive analytics, including numerical data summarization and visual analytics. Chapters five through seven discuss set theory, definitions and counting rules, probability, random variables, and probability distributions, with a number of business scenario examples. These chapters lay down the foundation for predictive analytics and model building. Chapter eight deals with statistical inference and discusses the most common testing procedures. Chapters nine through twelve deal entirely with predictive analytics. The chapter on regression is quite extensive, dealing with model development and model complexity from a user’s perspective. A short chapter on tree-based methods puts forth the main application areas succinctly. The chapter on data mining is a good introduction to the most common machine learning algorithms. The last chapter highlights the role of different time series models in analytics. In all the chapters, the authors showcase a number of examples and case studies and provide guidelines to users in the analytics field.

What Is Analytics?The Emergence and Application of AnalyticsSimilarities with and Dissimilarities from Classical Statistical AnalysisTheory versus Computational PowerFact versus Knowledge: Report versus PredictionActionable InsightSuggested Further ReadingIntroducing R—An Analytics SoftwareBasic System of RReading, Writing, and Extracting Data in RStatistics in RGraphics in RFurther Notes about RSuggested Further ReadingReporting DataWhat Is Data?Types of DataData Collection and PresentationReporting Current StatusMeasures of Association for Categorical VariablesSuggested Further ReadingStatistical Graphics and Visual AnalyticsUnivariate and Bivariate VisualizationMultivariate VisualizationMapping TechniquesScopes and Challenges of VisualizationSuggested Further ReadingProbabilityBasic Set TheoryThe Classical Definition of ProbabilityCounting RulesAxiomatic Definition of ProbabilityConditional Probability and IndependenceThe Bayes TheoremComprehensive ExampleAppendixSuggested Further ReadingRandom Variables and Probability DistributionsDiscrete and Continuous Random VariablesSome Special Discrete DistributionsDistribution FunctionsBivariate and Multivariate DistributionsExpectationAppendixSuggested Further ReadingContinuous Random VariablesThe PDF and the CDFSpecial Continuous DistributionsExpectationThe Normal DistributionContinuous Bivariate DistributionsIndependenceThe Bivariate Normal DistributionSampling DistributionsThe Central Limit TheoremSampling Distributions Arising from the NormalRandom Samples from Two Independent Normal DistributionsNormal Q-Q PlotsSummaryAppendixSuggested Further ReadingStatistical InferenceInference about a Single MeanSingle Population Mean with Unknown VarianceTwo Sample t-test: Independent SamplesTwo Sample t-test: Dependent (Paired) SamplesAnalysis of VarianceChi-Square TestsInference about ProportionsAppendixSuggested Further ReadingRegression for Predictive Model BuildingSimple Linear RegressionMultiple Linear RegressionANOVA for Multiple Linear RegressionHypotheses of Interest in Multiple Linear RegressionInteractionRegression DiagnosticsRegression Model BuildingOther Regression TechniquesLogistic RegressionInterpreting Logistic Regression ModelInterpretation and Inference for Logistic RegressionGoodness of Fit for the Logistic Regression ModelHosmer-Lemeshow StatisticsClassification Table and ROC CurveSuggested Further ReadingDecision TreesAlgorithm for Tree-Based MethodsImpurity MeasuresPruning a TreeAggregation Method: BaggingRandom ForestVariable ImportanceDecision Tree and Interaction among PredictorsSuggested Further ReadingData Mining and Multivariate MethodsDimension Reduction Technique: Principal Component AnalysisFactor AnalysisClassification ProblemDiscriminant AnalysisClustering ProblemSuggested Further ReadingModeling Time Series Data for ForecastingCharacteristics and Components of Time Series DataTime Series DecompositionAutoregression ModelsForecasting Time Series DataOther Time SeriesSuggested Further Reading

Ayanendranath Basu earned his PhD in statistics from The Pennsylvania State University in 1991, under the guidance of late Professor Bruce. G. Lindsay. After spending four years at the Department of Mathematics, University of Texas at Austin, as an assistant professor, he joined the Indian Statistical Institute in 1995. Currently, Dr. Basu is a professor of the Interdisciplinary Statistical Research Unit (ISRU), ISI-Kolkata. His research interests lie mainly in the following areas: minimum distance inference, robust inference, multivariate analysis, and biostatistics. Srabashi Basu earned her PhD in statistics from The Pennsylvania State University in 1992. After spending several years in University of Texas Health Science Center in San Antonio, she joined Indian Statistical Institute in 1995. Since 2006, Dr. Basu is working as an analytics specialist and independent consultant. She has extensive applied research publications to her credit. She also works as a corporate trainer in various areas of predictive analytics and machine learning. Dr. Basu has been an online instructor for Penn State Statistics World Campus courses since 2009. She also has developed online course materials in statistics, business analytics, R, and SAS.