Business Analytics with R and Python (AI for Risks)

Preface
Contents
1 Data Mining in Business
	1.1 Introduction
	1.2 Requirements for Data Mining
	1.3 Business Data Mining
		1.3.1 Frequent Itemset Mining
		1.3.2 Customer Relationship Management
		1.3.3 Bankruptcy Prediction
		1.3.4 Fraud Detection
	1.4 Summary
2 Data Mining Processes
	2.1 KDD
	2.2 CRISP-DM
		2.2.1 Business Understanding
		2.2.2 Data Understanding
		2.2.3 Data Preparation
		2.2.4 Modeling
		2.2.5 Evaluation
		2.2.6 Deployment
	2.3 SEMMA
		2.3.1 Step 1 (Sample)
		2.3.2 Step 2 (Explore)
		2.3.3 Step 3 (Modify)
		2.3.4 Step 4 (Model)
		2.3.5 Step 5 (Assess)
	2.4 Model Controls
	2.5 Evaluation of Model Results
		2.5.1 Example Model
		2.5.2 Cost Metrics
		2.5.3 Other Measures
	2.6 Summary
	References
3 Data Mining Software
	3.1 R
	3.2 Rattle
	3.3 Python
		3.3.1 Installing Python
		3.3.2 Running Python
	3.4 Summary
4 Association Rules
	4.1 Methodology
	4.2 Demonstration Dataset
		4.2.1 Fit
		4.2.2 Lift
	4.3 The Apriori Algorithm
	4.4 Association Rules from Software
		4.4.1 Association Rules in Rattle
		4.4.2 R Code
		4.4.3 Python Code
	4.5 Conclusion
	Reference
5 Cluster Analysis
	5.1 K-Means Clustering
		5.1.1 A Clustering Algorithm
		5.1.2 Loan Data
	5.2 Clustering Methods Used in Software
	5.3 Example Cases
		5.3.1 Churn Clustering Model
		5.3.2 Credit Risk Assessment Model
	5.4 Software
		5.4.1 Portuguese Bankruptcy Dataset
		5.4.2 Rattle K-Means Clustering
		5.4.3 R Clustering
		5.4.4 R Code
		5.4.5 Python Clustering
	5.5 File BostonHousingKaggle.csv
		5.5.1 R Code
	5.6 Summary
	References
6 Regression Algorithms in Data Mining
	6.1 Regression Models
	6.2 Forecasting S&P 500
		6.2.1 R Code for Simple Regression
		6.2.2 Python Code for Simple Regression
	6.3 ARIMA Modeling
		6.3.1 R Code for ARIMA
		6.3.2 Python Code for ARIMA
	6.4 Multiple Regression
		6.4.1 R Code for Multiple Regression
		6.4.2 Python Code for Multiple Regression
	6.5 Stepwise Regression
		6.5.1 R Code for Stepwise Regression
	6.6 Logistic Regression
		6.6.1 Tests of the Regression Model
		6.6.2 Software Demonstrations of Logistic Regression
		6.6.3 R Code
		6.6.4 Python Code
	6.7 Summary
7 Classification Tools
	7.1 Classification Models
		7.1.1 Regression
		7.1.2 Decision Trees
		7.1.3 Random Forest
		7.1.4 Extreme Boosting
		7.1.5 Support Vector Machines
		7.1.6 Neural Networks
	7.2 Bankruptcy Data Set
	7.3 Logistic Regression
		7.3.1 R Code Logistic Regression
		7.3.2 Python Code Logistic Regression
	7.4 Support Vector Machines
		7.4.1 R Code SVM
		7.4.2 Python Code SVM
	7.5 Neural Networks
		7.5.1 R Code Neural Network
		7.5.2 Python Code Neural Network
	7.6 Decision Trees
		7.6.1 R Code Decision Tree
		7.6.2 Python Code Decision Tree
	7.7 Random Forests
		7.7.1 R Code Random Forest
		7.7.2 Python Code Random Forest
	7.8 Boosting
		7.8.1 R Code XGBoost
		7.8.2 Python Code Gradient Boosting
	7.9 Comparison
	7.10 Loan Default Prediction Model
	7.11 Summary
	References
8 Variable Selection
	8.1 Taiwan Bankruptcy Data
		8.1.1 Correlation
		8.1.2 Logistic Regression Variable Significance
		8.1.3 Entropy
		8.1.4 Information Content from Random Forest Models
		8.1.5 Control Models Using All 94 Variables
	8.2 Example Variable Selection Case
	8.3 Value of Variable Reduction
	References
9 Dataset Balancing
	9.1 Bankruptcy Datasets
	9.2 Balancing
	9.3 Process
	9.4 Data
		9.4.1 Poland Data
		9.4.2 Taiwan Data
		9.4.3 Slovak Data
		9.4.4 U.S. Data
	9.5 Results
	9.6 Example Credit Card Fraud Detection Case
	9.7 Conclusions
	References
Index