This project was done as a Final Year Project (FYP) for my B.Eng. (Computer Engineering) under CG4003. Below is the abstract from the final report. I was supervised by Dr Akshay Narayan.

This work aims to investigate the use of Multi-Agent Reinforcement Learning (MARL) for coordinated problem solving by exploring the effectiveness of various algorithms. MARL is a subfield of Machine Learning where multiple agents learn to solve a problem together, that involves using RL concepts (Sutton & Barto, 2018), such as the Markov Decision Process (MDP), value functions, reward functions, and learning from environments. The work will focus on developing and evaluating different MARL algorithms that communicate (Comm-MARL) with the aim of knowledge sharing and coordination in various environments, implementing those MARL algorithms, and testing their performance on a common benchmark. The findings will contribute to the benchmarking of a Comm-MARL algorithm in an established MARL framework, and can provide a stepping stone to further development.

Knewbie was done as a summer project under Orbital, hosted by National University of Singapore (NUS) and supported by the likes of Google and PayPal. The platform was built having PSLE Mathematics students in mind, employing machine learning to provide tailored educational content to students for optimal learning. It is built using Python and the Flask framework.

From a total of 383 projects, Knewbie was 1 of 6 projects selected by the School of Computing (SoC) communication office, NUS and was featured on the various social media accounts.

Using Verilog, we built a Sound & Entertainment System on a Basys3 board with an OLED display and a microphone attached. We programmed Pong, Tetris, and responsiveness to audio for an entertaining and interactive experience.

As a Research Intern, I investigated Reinforcement Learning to optimize the Smart Control Strategy of Traffic Signals using SUMO and Python.

This internship was done as part of requirements for CG4003, alongside the FYP, and was supervised by Dr Zhang Yunchang.

As an SEP intern, I spearheaded use of modern NLP and ML technologies to provide valuable insight on news sentiment and collaborated with fellow peers to ensure reliability of site maintenance.