Machine Learning & Data Mining | CSE 450

03 Prepare : Reading

Preparation Material

Before coming to class this week, please work through the the following:

1. Confusion Matrices

   • DataSchool: Simple guide to confusion matrix terminology

2. Cross Validation

   • Udacity Video: K-Fold Cross Validation

   • sklearn documentation (Read at least the intro and 3.1.1)

3. Handling Data on Different Scales

   • About Feature Scaling and Normalization (Don't worry about the details of Naive Bayes or PCA, we'll talk about them in future weeks.)

4. Handling Categorical Data

   • Guide to Encoding Categorical Values in Python

5. Handling Missing Data in Python

   • Machine Learning Mastery

   • Data Science Handbook

6. Handling Missing Data in General

   Now that you have a basic idea of working with missing data in Python, we need to consider the proper strategies of what to do at a higher level.

   • First, we need to understand why it might be missing: Wikipedia and Measuring U: 3 Types of Missing Data

   • Now, what are some options: MeasuringU: 7 Ways to Handle Missing Data

Other Helpful Resources

• Working with Pandas - Data Camp: DataFrames in Python

• Pivoting, Stacking, and Unstacking - Reshaping in Pandas