purple sci-fi banner for Adi's GitHub profile

I'm a Graduate student and researcher at @MSU

I am currently building uncertainty-aware machine learning systems for noisy physical environments, places where standard confidence estimates break down and you need guarantees that actually hold.

My current work applies conformal prediction to gravitational wave detection, focusing on the gap between synthetic training data and real LIGO detector noise. Prior work uncovered the Evaluation Paradox in physics-constrained time series, proving that standard random splits mask temporal dependencies and produce invalid model selection.

I have a weird curiosity towards applying ML to world's most complex scientific problems and using programming to build that research infrastructure. I love studying Math, Statistics and Physics and doing nightlong coding sessions on my couch. Quantative Research also happens to be one of the fields I have an infatuation towards.

Outside of research, I like playing Football, Urban Exploration and reading Philosophy in my free time, if I am not watching "Suits".

Merci d'avoir lu.

• Physics-Aware Soil Moisture ML — ANN vs LSTM soil moisture prediction with uncertainty quantification and explainability
• Kerr Black Hole Simulator — physics-inspired black hole visualization engine using Three.js, WebGL, GLSL, and TensorFlow.js
• DB2-Platform — DB2 is a modern genome and transcriptomics platform for Onthophagus taurus dung beetle genome
• Hub-ITS-Triage-Agent — Call response system that can redirect calls to appropriate agencies and divisions across government

• Research/building in public on X 🔭
• Connecting professionally on LinkedIn 💼
• Sharing projects here on GitHub 🧑