1. Feature Extraction Portability

Feature extraction is highly dependent on the skill of analyst and is less studied with the coming of Deep Learning. However, they are still very important to achieve a goal.

2. Data Cleaning

• Standardization

  

• Min-Max Scaling

  

3. Data Reduction and Transformation

Dimensionality Reduction with axis rotation

• Principal Component Analysis: PCA
• Singular Value Decomposition: SVD
• CUR decomposition