I am a full-time undergraduate at the Singapore Institute of Technology (SIT) and will graduate with a Bachelor of Science (Honours) in Applied Artificial Intelligence in 2025. I was recognised for my academic achievements on the Provost's List for AY2022/23 for a YGPA of 4.75.
In 2019, I worked on predicting the onset of sepsis using the PhysioNet/CinC 2019 challenge dataset. My results were published in Computers in Biology and Medicine in 2020.
After graduation, I intend to pursue further studies in Artificial Intelligence, with a minor focus in the biomedical field if possible.
Over the course of the last 8 years, I have worked on a couple of projects that I am particularly proud of.
This project focused on predicting the onset of sepsis by 6 hours as part of the PhysioNet/CinC 2019 challenge. Findings were published in Computers in Biology and Medicine.
This project was undertaken as a requirement of a Computer Vision and Deep Learning module when studying at SIT. We fine-tuned a pre-trained ResNet-50 model with 5-fold cross validation. We evaluated its predictions on the validation set, then evaluated its robustness towards adversarial attacks.
This project was also a requirement for an Embedded Systems Programming module at SIT. My areas of responsibility were documentation and implementing the navigation for the car. We implemented and used the A* search, floodfill, and depth first search algorithms to map only the reachable areas of the maze, navigate back to the start of the maze, then navigate to the desired endpoint of the maze.
This project was a requirement for a Data Engineering and Visualisation module at SIT. We trained a custom Temporal Convolutional Network (TCN) on the MIT-BIH Arrhythmia Database with 10-fold cross validation. The models' predictions on the validation sets were used to augment the dataset according to which portions of the ECG signal the models find to be relevant to their predictions.
This project was a requirement for an Edge Computing and Analytics module at SIT. We used 4x Raspberry Pi 400s to collect Wi-Fi, Bluetooth Low Energy, and Webcam images to be used as features for a Gaussian Process Regressor to estimate the crowd level in a canteen.
This project was a requirement for a Database Systems module at SIT. The project utilised NextJS for the frontend, FastAPI as the API server, and MySQL and MongoDB for the database services. We also used Docker to containerise the frontend and backend for easier deployment.
Let me know how to get back to you.