03 — Density Estimation

Methods for estimating the probability density function underlying observed data — either non-parametrically (KDE) or through mixture models.

Notes

• Density Estimation — kernel density estimation (KDE), Gaussian mixture models (GMM), EM algorithm, model selection
• Density Estimation — Implementation — KDE (scipy), GMM BIC/AIC selection, Isolation Forest, Local Outlier Factor with scikit-learn

Links

← 02 — Dimensionality Reduction → 04 — Representation Learning

• Probability Distributions
• Bayesian Modeling — GMM as probabilistic model