Bias Variance Tradeoff
Reducing the Bias Variance Tradeoff and Over/Underfitting:
- Model Evaluation (Choosing the right model) KNN Tree Linear SVM
- Get More Data / Outlier handling
- Validation / Cross Validation (Test / train split) Only to be used with diagnosis, not to optimize your model.
- Domain Knowledge
- Feature Selection / Engineering
- Hyper Parameter Tuning (Regularization)