Collaborative Filtering in Recommendation Systems
Recommendation engines are one of the most widely used applications of machine learning. For example we can imagine I’m browsing Amazon for some new shoes. Perhaps a new sale has recently gone up from a seller who has previously sold to buyers similar to me – so Amazon chooses to recommend this sale. However this doesn’t take into account the value of the seller – for instance maybe the seller has very little history and I only want to buy from sellers that I deem trustworthy”. It also ignores whether the buyers who are similar to me make truly reliable recommendations. My project focuses on developing a more accurate recommendation engine that focuses in on different user and item intrinsic values and using these to minimize regret over recommendations for all users in the network.”

I’m a rising senior majoring in computer science and mathematics. I’ve always loved research – for two years of high school I did research a local university in computational biology. I love learning about the newest work in the field and working with the people who make it happen. I’m excited to spend the next year learning more about machine learning and I hope that my work will contribute to the field as a whole.