Projects
Some cool stuff that I have done!
Financial Coconut Transcripts
End-to-end ML project - data collection, transcription, summarisation, front-end website building and deployment - with MLOps to automate all that
Learn more...
Hierarchical Kalman Variational AutoEncoder for Video Prediction
Developed efficient method for video prediction for Masters Dissertation. Combined Variational AutoEncoders with Kalman Filters.
Learn more...
How Well Do Self-Supervised Models Transfer to Medical Imaging?
Published paper as part of Masters group project. Investigated the performance of pretrained SSL models in medical imaging.
Learn more...
Causal ML to infer heterogeneity in experiments
Best Undergraduate Dissertation. Employed causal ML to identify new heterogeneous effects for experiments previously analysed by OLS.
Learn more...
R package to compare between synthetic datasets
Contributions as part of Singapore Department of Statistics internship: First open-source package to systematically evaluate synthetic data based on differential privacy and accuracy of data.
Learn more...
Tabular Reinforcement Learning from Scratch
Implement and train RL algorithms (e.g. value iteration, dynamic programming, Monte Carlo & SARSA) from scratch.
Learn more...
Deep Reinforcement Learning from Scratch
Implement and train deep RL (DQN,DDQN) from scratch, including features like replay buffer, target network, and frame-skipping.
Learn more...
AI agent for Battleship
Wrote Battleship in Python. Programmed player to always win based on probabilities.
Learn more...
AI agent for Connect 4
Wrote Connect 4 in Python, and implemented alpha-beta pruning to search for optimal solution.
Learn more...