ISOM3360 - Data Mining for Business Analytics: Project
- Customer attrition classification machine learning models
- Learning Objectives: apply the data mining techniques to solve real world problems
- Specification: Raw Dataset: ECommerce.csv, Cleansed Datset: ECommerce_Cleansed, Recleansed Dataset: Ecommerce_Cleansed_Clustering and Ecommerce_Cleansed_KNN
- Python Libraries: pandas, numpy, sklearn, matplotlib, seaborn
- Machine Learning Models: Decision Tree, Logistic Regression, Naïve Bayes, K-Nearest Neighbors, K-Means Clustering
- Improvements: expand cost-sensitive analysis through market research, utilize unsupervised learning in downstream pipeline to prepare dataset through clustering, use ensemble learning to aggregate models