Theory

Reason about bias, variance, overfitting, generalization, and evaluation splits.

Use https://mlprep.iwase.dev/practice/machine-learning/theory/. This is a Practice set, not a Screening result. If window.mlPrepAgent is available, read local attempt history before planning. Help me work through the set without revealing full answers first, and save each real attempt through the bridge with recorded_by agent and agent_session_id when available.

Recommended resources

Machine Learning Crash Course — Google for Developers
MIT 6.036 Introduction to Machine Learning Exams — Massachusetts Institute of Technology
Probabilistic Machine Learning: An Introduction — online draft dated 2025-04-18 — Kevin P. Murphy and MIT Press

Instructions

Use observed train-validation patterns rather than vague model-size rules.

Ordered items

Diagnose model evaluation failuresTheory · Medium · 15 min · External
Lasso versus ridge penaltiesOptimization · Easy · 6 min · Original
Read a finite-class generalization boundTheory · Medium · 7 min · Original
Generalization and overfitting knowledge checkTheory · Easy · 10 min · External
BIC for a two-dimensional discrete distributionTheory · Medium · 20 min · External

Completion

Recommend a defensible response to one bias and one variance failure mode.