How Bayes' theorem helped win the second world war
Part I: Introduction to Bayesian statistics and how it cracked the Japaneses naval cipher JN 25
Portfolio
An overview of my academic research and data science projects.
Data Science & Machine Learning
NeuralXC
A Python software package that uses deep learning and symmetry preserving neural networks to enhance existing methods for simulating molecular systems, making them faster and more accurate.
Skills: deep learning, software development, TensorFlow, PyTorch
paper-scraper
Web-app that lets users explore recent arXiv publications based on their research interests.
Skills: AWS, Docker, dash/plotly, spectral-clustering, recommender systems, gensim
Twitter Sentiment Tracker
Web-app to track twitter users' sentiments towards topics and gain insights into market and cultural trends
Skills: AWS, Apache Spark, Databricks, ETL, streaming data, sentiment analysis, dash/plotly
So you want to move to...?
Finding (Pareto) optimal neighborhoods in NYC
Skills: (polynomial, k-NN, random forest)-regression, data visualization, scikit-learn, dash/plotly
Saturation spectroscopy
Data analysis workflow of a saturation spectroscopy study on Rubidium. The experiment was conducted as part of the Graduate Laboratory course at Stony Brook University.
Skills: scientific data analysis, dashboards
Political twitter
Analysis of the US political twitter landscape.
Skills: ETL, sentiment analysis, hypotheses testing, data visualization, bokeh, networkx
Writing
Supercharge your model performance with inductive bias
How to work smarter, not harder: Encoding real-life symmetries in machine learning models can improve their accuracy by orders of magnitude!
Abstract base classes and how to use them in your data science project
Write cleaner and safer python code by using this object-oriented programming building block
Unwrapping the Swiss Roll with Diffusion Maps
The amazing dimensionality reduction algorithm you may have never heard of