Avatar

Yihan Zhou(Joey)

Master's Student

The University of British Columbia

About Me

I'm a second year Master's student at UBC, fortunately co-supervised by Nick Harvey and Mark Schmidt. My research revolves around optimization in gradient descent methods and online learning. In addition, I'm interested in learning theory and more generally theoretical computer science.

I enjoy poems, movies and literature. García Márquez is my favorite writer. I watch soccer, basketball and Dota. Liverpool, Celtics and VG are my home teams.

I'm one with the force and the force is with me!

Interests

  • Optimization in ML
  • Online Learning
  • Learning Theory

Education

  • MSc in Compter Science, 2018-Ongoing

    The University of British Columbia

  • BMath in CS and joint C&O, 2015-2018

    University of Waterloo

Projects

*

Machine Learning Algorithms for White Balancing

Developing ML algorithms for white balancing in order to achieve better performance.

Privacy in Machine Learning

A survey on privacy in machine learning.

Private Sequential Learning

Extending current theoretical results on private sequential learning to Bayesian setting and higher dimensions.

Replicating and Improving the Prediction of Antidepressant Response Using the CAN-BIND and STAR*D Datasets

Replicating a study using ML algorithms for the prediction of antidepressant response on a different dataset and attempting to improve …

Variants of Frank-Wolfe algorithm

Studying Frank-Wolfe algorithm and its variants.

Recent Talks

Generalization of Kernel Learning

Similar to deep neural networks, kernel machines have good generalization behaviour. However, such behaviour cannot be explained by …

Online Learning and Bandits

An overview of online learning and different bandit problem.

Dictionary Learning

An introduction to dictionary learning and its applications.

Working Experience

 
 
 
 
 

Teaching Assistant

UBC

Sep 2018 – Present Vancouver
TA for third/fourth year algorithm and ML courses.
 
 
 
 
 

Software Developer

NN Life Japan

Sep 2016 – Dec 2016 Tokyo
Designing and implementing a daily process in Windows PowerShell to fetch useful data and use the data to make predictions on the number of expected telephone calls daily. Using C# and Visual Studio to develop tools with simple WPF UI in order to simplify certain operations.
 
 
 
 
 

Web Developer

Qlik

Jan 2016 – Apr 2016 Ottawa
Building a web app named Pythia to extract data from Google Analytics, New Relic, AWS EC2 and Elastic Search by making REST API calls and upload such data to Mongo Database. Creating an automatically updated dashboard of product information to improve team working efficiency.