BIOMEDIN 273B: Deep Learning in Genomics and Biomedicine

Time: 2020 fall

Description: This course was taught by professor Anshul Kundaje and James Zou. In the course, we formed groups and were given data and directions for course projects. Our group had 4 people, with data from an HT-recruit RNA-seq study, we wanted to build models to identify repression effect of protein tiles and Pfam proteins.

Responsibility: I was responsible for building coding pipeline for data loading, visualization of figures, model tuning and prediction analysis, finalizing the presentations and final report.

Code: The code of this project can be found here.

Report: final report of the project is recorded below.

CS 330: Deep Multi-Task and Meta Learning

Time: 2020 fall

Description: This course was taught by professor Chelsea Finn, and was focus on meta-learning and multi-task learning subjects, especially in reinforced-learning topics. The course project is freely selected topic/dat with course related research ideas.

Project Overview: I used head impact signal data from multiple sources/sport type/datasets, which maximum principal strain is a important prediction target. By implementing meta-learning approaches to do the prediction, I hoped to let the models to be adaptive for multiple datasets. Data distribution shifting as well as long-tailed distribution datasets are very common in medical field (e.g. institutional differences of medical images, rare disease, patient cohort changes)

Code: This project is still under development. I want to try some new ideas, so the code is not organized yet.

Report: final report of the project is recorded below.

CS 230: Deep Learning

Time: 2019 winter

Description: Taught by the CS230 teaching team led by Andrew Ng, focused on general foundations of deep learning. Open topic project.

Poster/Project Overview: Poster of the project can be accessed through this link.

Report: Final report of the project can be accessed through this link.

CS 229: Machine Learning

Time: 2020 spring

Description: Taught by professor Tengyu Ma, Andrew Ng and Chris Ré. The “most famous” course in Stanford. Open topic project.

Poster/Project Overview: Poster of the project can be accessed through this link.

Report: Final report of the project can be accessed through this link.

BIOMEDIN 260: [Computational Methods for Biomedical Image Analysis and Interpretation](https://explorecourses.stanford.edu/search?view=catalog&filter-coursestatus-Active=on&q=BIOMEDIN 260: Computational Methods for Biomedical Image Analysis and Interpretation&academicYear=20142015)

Time: 2020 spring

Description: Taught by professor Daniel L. Rubin, mainly about using deep models to deal with medical images.

Project Overview: In a group of two people, we went for a kaggle competition dataset with a x-ray dataset of pneumothorax. The goal was to built a efficient classification model to distinguish patients with pneumothorax from those who didn’t. The majority of the project focused on using transfer learning, by fine-tuning models based on state-of-the-art 2D image nets and compare the results.

Report: final report of the project is recorded below.