Hi, I'm Zach. I love artificial intelligence, data science, and computer science. Look below to learn more about me and see my projects!
>>> Zach.location
'Berkeley, CA'

>>> Zach.education
{'school' : 'Univeristy of California, Berkeley', 'gpa' : 3.7, 'expectedGraduation' : 'May 2020', "major" : "Computer Science"}

>>> Zach.interests
'["Artificial Intelligence", "Data Science", "Efficient Algorithms", "Weightlifting", "Triathalons", "Entrepreneurship"]'

>>> Zach.email
'zachchao@berkeley.edu'

>>> Zach.resume
resume.pdf

>>>

Projects

Card image cap
Customer Complaint Analysis

Won second in GM's Machine Learning hackathon. Created a custom word embedding model which mapped customer complaints to subareas. Then mapped each subarea to the engineering defect which was most likely to have caused it the complaints within that subarea.

Card image cap
Reinforcement Learning

Different Reinforcement Learning algorithms from David Silver's lectures at University College London applied to OpenAI's gym and other games.

Card image cap
Expectimax Game AI

Won fourth out of six hundred in Berkeley competition for best AI to solve a complicated game with up to ten dice. Implemented expectimax using dynamic programming and probability theory.

Card image cap
Web Crawlers

Much of the work in data science is data acquisition and cleaning. Created a multitude of spiders using Python requests and scrapy to collect labeled data and pipeline it databases for machine learning and data science.

Card image cap
Group Project Designs

Created group projects in MiraCosta college's Data Structures and Algorithms class. Served as SCRUM master on the projects to teach students SCRUM, linked lists, MVC, Git, UML and JUnit testing.

Card image cap
Hashtag Sentiment Analysis

Using word2vec, TSNE and KNN, created a model which takes a hashtag and outputs a more popular related hashtag which caters to a larger audience, optimizing usage.

Card image cap
Instagram Data Analytics

A service which, when given an Instagram username will fetch the user's data and using several machine learning models and data analytics, outputs improvements the user can make as well as the percent improvement to be expected.