Research materials

These are the materials that I find either:

• useful, or at least deemed to be useful;
• hard to understand;
• fun to work with, or
• all of the above.

2019

Machine Learning

• Cross-validation for fusion estimates
• Imputation using EM: Python implementation

Methods for Multivariate Data

• Canonical correlation analysis
• Imputing missing data using EM algorithm
• Justifying matrix derivatives
• Linear regression with multivariate response
• PCA on correlation matrix

Stochastic Processes

• Animation in R: a jumping frog
• Animation in R: the largest value until the current roll

2018

Methods of Applied Statistics

• Binarizing data using data.table in R
• Censored data
• Linear mixed effects model
• Visualizing confidence regions

Ordinary Differential Equations

• Visualizing Mankiw-Romer-Weil growth model

Statistical Computation

• Boundary estimation
• Dealing with a non-separable penalty term
• Evaluating a hard-to-evaluate pmf using pgf and DFT
• Using pgf and DFT: Python implementation

Time Series Analysis

• Study note 1: weak stationarity, ARIMA, Dickey-Fuller
• Study note 2: TFNM
• Study note 3: VAR, Granger causality, cointegration, multivariate times series
• Study note 4: state-space model, bootstrap
• Study note 5: problems, summary