IBM SPSS Modeler Tutorial on Data Mining Process
In this video tutorial, I cover the main features of IBM SPSS Modeler so that you can run an end-to-end data mining process. I use a sample data to demonstrate all the required steps in action to build both predictive (supervised learning) and descriptive (unsupervised learning) models, and provide further explanations and tips on the way.
This tutorial was originally made for students taking the ADM3308 (Business Data Mining) course at the University of Ottawa in winter 2020, and covers the following topics:
1. Introduction to IBM SPSS Modeler
2. Import source data (CSV, Excel)
3. Data exploration
4. Identify and impute missing values
5. Data cleaning (fix noises)
6. Binning or categorization
7. Data partitioning or splitting
8. Data balancing (undersampling, oversampling, SMOTE)
9. Decision tree: C5
10. Neural network: MLP
11. Clustering: K-Means
12. Confusion matrix and accuracy
13. AUC and ROC curve
14. Visualization graphs
15. Model deployment and export