Advanced Modeling in Data Mining
CHAPTER 1: PREPARING DATA FOR MODELING
CLEANING DATA
BALANCING DATA
NUMERIC DATA TRANSFORMATIONS
BINNING DATA VALUES
DATA PARTITIONING
ANOMALY DETECTION
FEATURE SELECTION FOR MODELS
CHAPTER 2: DATA REDUCTION: PRINCIPAL COMPONENTS
USE OF PRINCIPAL COMPONENTS FOR PREDICTION MODELING AND CLUSTER ANALYSES
WHAT TO LOOK FOR WHEN RUNNING PRINCIPAL COMPONENTS OR FACTOR ANALYSIS
PRINCIPLES
FACTOR ANALYSIS VERSUS PRINCIPAL COMPONENTS ANALYSIS
NUMBER OF COMPONENTS
ROTATIONS
COMPONENT SCORES
SAMPLE SIZE
METHODS
OVERALL RECOMMENDATIONS
EXAMPLE: REGRESSION WITH PRINCIPAL COMPONENTS
CHAPTER 3: DECISION TREES/RULE INDUCTION
COMPARISON OF DECISION TREE MODELS
USING THE C NODE
BROWSING THE MODEL
GENERATING AND BROWSING A RULE SET
UNDERSTANDING THE RULE AND DETERMINING ACCURACY
UNDERSTANDING THE MOST IMPORTANT FACTORS IN PREDICTION
FURTHER TOPICS ON C MODELING
MODELING SYMBOLIC OUTPUTS WITH OTHER DECISION TREE ALGORITHMS
MODELING SYMBOLIC OUTPUTS WITH CHAID
MODELING SYMBOLIC OUTPUTS WITH C&R TREE
MODELING SYMBOLIC OUTPUTS WITH QUEST
INTERACTIVE TREES
PREDICTING NUMERIC FIELDS
CHAPTER 4: NEURAL NETWORKS
TO NEURAL NETWORKS
THE NEURAL NETWORK NODE
MODELS PALETTE
VALIDATING THE LIST OF PREDICTORS
UNDERSTANDING THE NEURAL NETWORK
UNDERSTANDING THE REASONING BEHIND THE PREDICTIONS
USING AN ANALYSIS NODE
MODEL SUMMARY
ADVANCED NEURAL NETWORK TECHNIQUES
TRAINING METHODS
THE MULTILAYER PERCEPTRON
THE RADIAL BASIS FUNCTION
EXPERT OPTIONS
AVAILABLE ALGORITHMS
WHICH METHOD, WHEN?
PREVENTION OF OVERTRAINING
MISSING VALUES IN NEURAL NETWORKS
EXPLORING THE DIFFERENT NEURAL NETWORK OPTIONS
CHAPTER 5: SUPPORT VECTOR MACHINES
THE STRUCTURE OF SVM MODELS
SVM MODEL TO PREDICT CHURN
EXPLORING THE MODEL
A MODEL WITH A DIFFERENT KERNEL FUNCTION
TUNING THE RBF MODEL
CHAPTER 6: LINEAR REGRESSION
BASIC CONCEPTS OF REGRESSION
AN EXAMPLE: ERROR OR FRAUD DETECTION IN CLAIMS
CHAPTER 7: COX REGRESSION FOR SURVIVAL DATA
WHAT IS SURVIVAL ANALYSIS?
COX REGRESSION
COX REGRESSION TO PREDICT CHURN
CHECKING THE PROPORTIONAL HAZARDS ASSUMPTION
PREDICTIONS FROM A COX MODEL
CHAPTER 8: TIME SERIES ANALYSIS
WHAT IS A TIME SERIES?
A TIME SERIES DATA FILE
TREND, SEASONAL AND CYCLIC COMPONENTS
WHAT IS A TIME SERIES MODEL?
INTERVENTIONS
EXPONENTIAL SMOOTHING
ARIMA
DATA REQUIREMENTS
AUTOMATIC FORECASTING IN A PRODUCTION SETTING
FORECASTING BROADBAND USAGE IN SEVERAL MARKETS
APPLYING MODELS TO SEVERAL SERIES
CHAPTER 9: LOGISTIC REGRESSION
TO LOGISTIC REGRESSION
A MULTINOMIAL LOGISTIC ANALYSIS: PREDICTING CREDIT RISK
INTERPRETING COEFFICIENTS
CHAPTER10: DISCRIMINANT ANALYSIS
HOW DOES DISCRIMINANT ANALYSIS WORK?
THE DISCRIMINANT MODEL
HOW CASES ARE CLASSIFIED
ASSUMPTIONS OF DISCRIMINANT ANALYSIS
ANALYSIS TIPS
COMPARISON OF DISCRIMINANT AND LOGISTIC REGRESSION
AN EXAMPLE: DISCRIMINANT
CHAPTER 11: BAYESIAN NETWORKS
THE BASICS OF BAYESIAN NETWORKS
TYPE OF BAYESIAN NETWORKS IN CLEMENTINE
CREATING A BAYES NETWORK MODEL
MODIFYING BAYES NETWORK MODEL SETTINGS
CHAPTER 12: FINDING THE BEST MODEL FOR BINARY OUTCOMES
CHAPTER 13: FINDING THE BEST MODEL FOR NUMERIC OUTCOMES
CHAPTER 14: GETTING THE MOST FROM MODELS
COMBINING MODELS WITH THE ENSEMBLE NODE
USING PROPENSITY SCORES
METALEVEL MODELING
ERROR MODELING
