I'm a Computer Science student based in Melbourne, exploring artificial intelligence and artificial life. I try to develop systems that not only mimic life but also offer new perspectives on what it means to be 'intelligent'.
- Neural Network Generator Tool (TypeScript, React)
- This project is a practical response to the challenges I've noticed in designing neural networks, especially convolutional neural networks where small changes in parameters can drastically change the output shape and require a lot of trial and error or forethought. Using TypeScript, React, and Node.js, I've built this tool to make it easier for users to experiment with different neural network layers and parameters. It automatically calculate output shapes, ensuring that different layers work well together.
- Knights-Tour Neural Network Solver (Python)
- The Knights-Tour Neural Network Solver can handle arbitrary board sizes, provided an excessive amount of processing power. The tools was designed using only Numpy and pygame (for visualisation) without any standard libraries like PyTorch or TensorFlow. Neural networks are not an optimal tool for solving the Knights-Tour, but this was a fun experiment and challenge that helped me get a better grasp of the basics of neural networks.
- Tic-Tac-Toe Algorithms Project (Python, PyTorch)
- The Tic-Tac-Toe Algorithms Project features various strategies like heuristic, Minmax, and Monte-Carlo Exploring Starts, as well as Q-Learning, SARSA, Deep Q-Learning, and Actor-Critic methods. Algorithms can be selected and pitted against each other, making it a useful tool for exploring different AI techniques.