A guide to my miscellaneous projects.
- Explored the intersection of psychology and data science, showcasing a dramatic reduction in lines required to invoke an interactable map by 99% using my modular library, facilitating seamless exploration of their relationship, as demonstrated in this YouTube video.
- Boosted deep learning efficiency by 2x through HPC-optimized training of a CNN model, accelerating computation speed.
- Enhanced gaming AI by ensuring decision trees are reduced by a factor of log 2 in a TicTacToe AI with partial alpha-beta pruning.
- Developed an MNIST classifier in Python and C from scratch, with 15% faster classification than standard libraries.
- Created a handwritten digit predictor utilizing tensorflow.js, offering seamless prediction with a full-stack architecture, while reducing model footprint by 97%.