Machine Learning

Model assumptions, bias/variance, and generalization behavior

Losses, optimizers, regularization, and neural network training mechanics

Cross-validation, metrics, calibration, and validation design

Leakage, resampling, feature encoding, and preprocessing pitfalls

Linear models, regularization, bagging, boosting, and tree ensembles

Neural network layers, normalization, dropout, and residual paths

Feature attribution, counterfactuals, and explanation limits

Bayesian prediction, uncertainty estimates, and prediction sets

Deployment drift, data quality, and post-launch model health

Contrastive learning, self-supervision, and collapse-prevention objectives

VAEs, flows, diffusion models, and generative training objectives

Causal estimators, orthogonal scores, off-policy evaluation, and RL advantage estimates