Master’s Degree Project: Distributed Key-Value Database

---

● Implemented the specifications of a key-value database that is sharded and replicated across many servers.
● The db has no single point of failure because it uses Paxos to replicate data across the nodes of a shard.
● Not all nodes are responsible for all data; groups of nodes save certain shards of the data.
● Shard reorganization code allow nodes to join the db and cause the number of shards to increase.

• Date: December 2015

Close

Third Place Winning Team at a Hackathon in UC Berkeley

---

I worked on a flask web app that predicts future salaries of your LinkedIn connections and gives you recommendations on whom you should ask for job referrals. This project won 3rd place at a UC Berkeley hackathon that was hosted by Intuit.

I designed the strategy on applying linear regression models and cross validation to do salary predictions. I also implemented the machine learning code using Python and the Scikit-learn module.

• Date: September 2016
• Link to Project: https://github.com/andrewyli/hackintoit2016

Close

Vaccine Usage Predictor

---

This project involved building linear regression models with Python's Scikit-learn library to predict vaccination usage in Denmark. Cross validation was performed in order to reduce model over-fitting. The MMR vaccination predictor was accurate enough to achieve a root-mean-square-error that was less than 15% of the data range.

• Date: April 2016
• Link to Project: kenluy.com/vaccine.pdf

Close

My Personal Website

---

I created a website to show my projects and contact information. This website was made by using Bootstrap and hosting it on GitHub Pages.

• Start Date: December 2016
• Link to Project: kenluy.com

Close