Machine Learning research is currently one of my main focuses.
Here are some of my developing papers, with more to be added over the coming months.
Utilized coreset selection methods such as KCenterGreedy and Active Learning to reduce the time it takes to fine tune an LLM, while retaining performance metrics (reducing training time by 80%). Placed second at ProjectX, the world's largest international undergraduate machine learning contest among leading universities.
View PaperMy team and I worked with a Kenyan university to solve the problem of long ATM lines in Kenya. We employed numerous strategies, such as Yolov8 and Deep CNN's to solve this problem.
View Paper