Web Analytics

Projects

I am interested in machine learning and data science. Below are some projects that I have worked on in the past.

2017 NHL Expansion Draft Optimizer

The Vegas Golden Knights entered the NHL in the 2017-2018 season. On June 21, 2017, the team drafted 30 players, one from each of the 30 current NHL teams, subject to a set of drafting and protection rules that all teams must follow. We developed a web app to simulate the drafting and protection problems in real-time given a set of management preferences such as maximizing the on-ice performance or financial flexibiltiy for teams.

Generative Adversarial Networks for Computationally Efficient Radiotherapy

Knowledge-based planning (KBP) prediction models estimate the dose of radiation to be delivered to cancer patients using patient features. We proposed a generative adversarial network (GAN) to predict the entire 3-D dose distribution from contoured CT images. Experiments on a clinical data set of head-and-neck cancer patients showed that our GAN-based approach significantly outperformed prior deep learning and classical machine learning methods when evaluated on clinical satisfaction criteria.

The Open Knowledge-Based Planning Challenge

Knowledge-based planning (KBP) is an automated approach to radiation therapy treatment planning that involves first predicting a desirable treatment plan, before correcting it to a deliverable one. Current research on automated planning is limited due to private clinical data sets and evaluation metrics. We created an open and standardized data set along with an OpenKBP Challenge to adavance fair and consistent comparisons of KBP prediction models. This challenge ran from January 2020 to June 2020.

2021 NHL Expansion Draft Optimizer

The Seattle Kraken entered the NHL in the 2021-2022 season by drafting 30 players from the current NHL teams on July 21, 2021, subject to a set of drafting and protection rules. We developed a web app to simulate the drafting and protection problems in real-time given a set of management preferences such as maximizing the on-ice performance or financial flexibiltiy for teams.