1st Edition

Computational Business Analytics

By Subrata Das Copyright 2014
    516 Pages 290 B/W Illustrations
    by Chapman & Hall

    Learn How to Properly Use the Latest Analytics Approaches in Your Organization

    Computational Business Analytics presents tools and techniques for descriptive, predictive, and prescriptive analytics applicable across multiple domains. Through many examples and challenging case studies from a variety of fields, practitioners easily see the connections to their own problems and can then formulate their own solution strategies.

    The book first covers core descriptive and inferential statistics for analytics. The author then enhances numerical statistical techniques with symbolic artificial intelligence (AI) and machine learning (ML) techniques for richer predictive and prescriptive analytics. With a special emphasis on methods that handle time and textual data, the text:

    • Enriches principal component and factor analyses with subspace methods, such as latent semantic analyses
    • Combines regression analyses with probabilistic graphical modeling, such as Bayesian networks
    • Extends autoregression and survival analysis techniques with the Kalman filter, hidden Markov models, and dynamic Bayesian networks
    • Embeds decision trees within influence diagrams
    • Augments nearest-neighbor and k-means clustering techniques with support vector machines and neural networks

    These approaches are not replacements of traditional statistics-based analytics; rather, in most cases, a generalized technique can be reduced to the underlying traditional base technique under very restrictive conditions. The book shows how these enriched techniques offer efficient solutions in areas, including customer segmentation, churn prediction, credit risk assessment, fraud detection, and advertising campaigns.

    Analytics Background and Architectures
    ANALYTICS DEFINED
    ANALYTICS MODELING
    ANALYTICS PROCESSES
    ANALYTICS AND DATA FUSION

    Mathematical and Statistical Preliminaries
    STATISTICS AND PROBABILITY THEORY
    LINEAR ALGEBRA FUNDAMENTALS
    MATHEMATICAL LOGIC
    GRAPHS AND TREES
    MEASURES OF PERFORMANCE
    ALGORITHMIC COMPLEXITY

    Statistics for Descriptive Analytics
    PROBABILITY DISTRIBUTIONS
    DISCRETE PROBABILITY DISTRIBUTIONS
    CONTINUOUS PROBABILITY DISTRIBUTIONS
    GOODNESS-OF-FIT TEST

    Bayesian Probability and Inference
    BAYESIAN INFERENCE
    PRIOR PROBABILITIES

    Inferential Statistics and Predictive Analytics
    CHISQUARE
    TEST OF INDEPENDENCE
    REGRESSION ANALYSES
    BAYESIAN LINEAR REGRESSION
    PRINCIPAL COMPONENT AND FACTOR ANALYSES
    SURVIVAL ANALYSIS
    AUTOREGRESSION MODELS

    Artificial Intelligence for Symbolic Analytics
    ANALYTICS AND UNCERTAINTIES
    NEO-LOGICIST APPROACH
    NEO-PROBABILIST
    NEO-CALCULIST APPROACH
    NEO-GRANULARIST

    Probabilistic Graphical Modeling
    NAIVE BAYESIAN CLASSIFIER (NBC)
    KDEPENDENCE NAIVE BAYESIAN CLASSIFIER (KNBC)
    BAYESIAN BELIEF NETWORKS

    Decision Support and Prescriptive Analytics
    EXPECTED UTILITY THEORY AND DECISION TREES
    INFLUENCE DIAGRAMS FOR DECISION SUPPORT
    SYMBOLIC ARGUMENTATION FOR DECISION SUPPORT

    Time Series Modeling and Forecasting
    PROBLEM MODELING
    KALMAN FILTER (KF)
    MARKOV MODELS
    DYNAMIC BAYESIAN NETWORKS (DBNS)

    Monte Carlo Simulation
    MONTE CARLO APPROXIMATION
    GIBBS SAMPLING
    METROPOLIS-HASTINGS ALGORITHM
    PARTICLE FILTER (PF)

    Cluster Analysis and Segmentation
    HIERARCHICAL CLUSTERING
    K-MEANS CLUSTERING
    K-NEAREST NEIGHBORS
    SUPPORT VECTOR MACHINES
    NEURAL NETWORKS

    Machine Learning for Analytics Models
    DECISION TREES
    LEARNING NAIVE BAYESIAN CLASSIFIERS
    LEARNING OF KNBC
    BAYESIAN BELIEF NETWORKS
    INDUCTIVE LOGIC PROGRAMMING

    Unstructured Data and Text Analytics
    INFORMATION STRUCTURING AND EXTRACTION
    BRIEF INTRODUCTION TO NLP
    TEXT CLASSIFICATION AND TOPIC EXTRACTION

    Semantic Web
    RESOURCE DESCRIPTION FRAMEWORK (RDF)
    DESCRIPTION LOGICS

    Analytics Tools
    INTELLIGENT DECISION AIDING SYSTEM (IDAS)
    ENVIRONMENT FOR FIFTH GENERATION APPLICATIONS (E5)
    ANALYSIS OF TEXT (ATEXT)
    R AND MATLAB
    SAS AND WEKA

    Analytics Case Studies
    RISK ASSESSMENT MODEL I3
    RISK ASSESSMENT IN INDIVIDUAL LENDING USING IDAS
    RISK ASSESSMENT IN COMMERCIAL LENDING USING E5 AND IDAS
    FRAUD DETECTION
    SENTIMENT ANALYSIS USING ATEXT

    Appendix A: Usage of Symbols
    Appendix B: Examples and Sample Data
    Appendix C: MATLAB and R Code Examples

    Index

    Further Reading appears at the end of each chapter.

    Biography

    Subrata Das is the founder and president of Machine Analytics and also serves as a consulting scientist to other companies. He has many years of experience in industrial, government, and academic research and development. He earned his Ph.D. in computer science and master's in mathematics.