Maze generator and solver written in Python. The project utilizes multiple algorithms to generate and solve mazes, including A* and Kruskal's algorithms.

Donald Knuth's Dancing Links algorithm written in C++. And one written in Python with PyGame to provide a visual representation of the algorithm.

2024-25 Fall Semester Software Engineering project for the course. This is a Finance Tracker app that is designed to help you track your spending changes or habits and analyze them.

This is my portfolio website that you are currently viewing. It is built with Next.js, React, and Tailwind CSS. It is a Full Stack application that showcases my projects and skills. It utilizes Firebase on the backend for authentication and database management.

There are a handful of ML projects that I have done over the years. Below is a list of the ones I am the most proud of.

CNN MNIST Digit Classifier

• Trained a CNN to classify handwritten digits (MNIST).
• Used PyTorch with convolution, pooling, and dense layers.
• Visualized confusion matrix and training loss.
• Evaluated using accuracy and F1 score.

CNN MNIST Digit Classifier

View Project

Face Clustering with FaceNet + HDBSCAN

• Face Detector: Multi-task Cascaded Convolutional Networks (MTCNN)
• Embedding Model: InceptionResnetV1 pretrained on VGGFace2
• Clustering Algorithm: Hierarchical DBSCAN (density-based, handles noise)
• Dimensionality Reduction: PCA for visualization only (clustering uses full embeddings)

Face Detection Project

View Project

Variational Autoencoder MNIST

• Implemented VAE in PyTorch.
• Learned latent space representation.
• Used KL divergence + reconstruction loss.
• Visualized latent embeddings.

VAE MNIST

View Project