Automatic Integration for Neural Spatio-Temporal Point Process models (AI-STPP) is a new paradigm for exact, efficient, non-parametric inference of point process. It is capable of learning complicated underlying intensity functions, like a damped sine wave.

My chapter summaries, example solutions, and implementation for the fantastic book "Pattern recognition and machine learning" by Christopher Bishop.

Spatial and temporal epidemiology data mining flow tools including data processing and analysis, model setup and simulation, inference and evaluation. Focusing on state-of-the-art methods such as universal differential equations, epidemiology-informed deep learning methods.

This repository contains code for paper: "Are you sure it’s an artifact? Artifact detection and uncertainty quantification in histological images."

Bridging System dynamics tool (Vensim) and Bayesian computation tool (Stan)

Linear models, Bayesian multivariate statistics, probabilistic machine leanring.

Framework for learning effective reduced order dynamics of molecular systems.

Code and thesis for my MSc dissertation on Bayesian unsupervised learning with missing data for mixture modeling. Implements Gibbs sampling and Variational Bayes EM for Gaussian and Bernoulli mixture models, with extensions to MNAR settings. Includes simulation pipelines, evaluation benchmarks, and the full dissertation document.

We use probabilistic programming (PyMC3) to roughly estimate 'R_0' and 'lambda_0' directly from pandemic infection data.

Citadel India Terminal 2025