How Dirichlet processes become clustering models through the CRP, stick-breaking, and mixture likelihoods.
StatisticsML
Notes on research, methods, and things I wanted to understand better. Search or filter by tag, or subscribe via RSS.
How Dirichlet processes become clustering models through the CRP, stick-breaking, and mixture likelihoods.
A mathematical introduction to directional data, Bessel-function asymptotics, and stable normalization on the hypersphere.
A reminder that simple ≠ weak.
A (long) introduction to neural nets, and popular options of CNNs, RNNs, Transformers, and other modern machine learning models
A brief review of my predoc applications during AY 2024–2025
Times New Roman, Palatino, Garamond, Erewhon, Georgia
Using pairwise comparison data to assess abilities
A quick guide to LaTeX