Result.md

Regression Models

• Scaling methods: None, StandardScaler, MinMaxScaler, MaxAbsScaler, RobustScaler, Normalizer
• Score method: r2, f1 (Logistic Regression)
• Score is calculated by:
  • Choose Scaling method then predict 100 times
  • Get the Mean value of 100 times prediction

  Score is the highest Mean of 100 times prediction of each Scaling methods

Logistic Regression

• Best Scaler: RobustScaler
• Score: 0.9

Support Vector Regression

• Best Scaler: MaxAbsScaler
• Score: 0.784

SGD Regression

• Best Scaler: MinMaxScaler
• Score: 0.776

Random Forest

• Best Scaler: None
• Score: 0.769

Ridge Regression

• Best Scaler: MinMaxScaler
• Score: 0.759

Gaussian Regression

• Best Scaler: MinMaxScaler
• Score: 0.759

K-Nearest Neighbors Regression

• Best Scaler: None
• Score: 0.757

Linear Regression

• Best Scaler: MinMaxScaler
• Score: 0.743

Huber Regression

• Best Scaler: MaxAbsScaler
• Score: 0.729

Lasso Regression

• Best Scaler: None
• Score: 0.63

Decision Tree Regression

• Best Scaler: MaxAbsScaler
• Score: 0.517