MLVU 4.2: Class imbalance and feature design @riskone1

Uploaded February 2021 | Updated September 2026, 3 days ago

Continuing our theme of pre-processing, we discuss how to deal with class imbalance, how to transform given data to features, and how to improve the expressive power of your model by adding new features based on the existing ones.

slides: mlvu.github.io/lectures/22.Methodology2.annotated.pdf
lecturer: Peter Bloem

MLVU 4.2: Class imbalance and feature design

12 Matrix models: Recommender systems, PCA and Graph convolutions

MLVU 1.5 Generalization in machine learning

MLVU 8.4: Expectation-maximization from first principles

10 Tree Models and Ensembles: Decision Trees, Boosting, Bagging, Gradient Boosting (MLVU2018)

12 Matrix Models: Recommender systems and Matrix factorization (MLVU2018)