Hi, there! I am Ye Xiao, currently a second year data science student at University of Southern California. I am a software developer with a keen interest in problem-solving, delivering scalable solutions and always prepared to exploring the new technologies.
I have experience building web application across Java, Python and Javascript. I also have good experience in data analysis/data visualization. My goal is to grow as a full stack developer, where ability to process data comes natural to me. Currently I'm open to all full-time opportunity, if you have interest or any idea to discuss, feel free to reach me out via the following email.
Ye Xiao
1253 W 35th St
Los Angeles, CA 90007 US
(213) 590-6149
seanxiao1996@gmail.com
Master in Data Science • Jan 2019 - Now
Relevant Coursework: (Algorithm:A, Machine learning:A-, Database:A-, Data Mining:A-, Data Visualization:A, Knowledge Graph:A)
B.S. Degree in Applied Math / Physics • Sept 2014 - Jun 2018
Software Engineer • Feb 2018 - Aug 2018
Implemented the resource alarm module of the web portal for FitOS(cloud operating system) via Django. Cooperated with the testing group to debug and optimize code performance. Designed the PRD for future refactoring of the project.
Software Engineer Internship • July 2017 - Aug 2017
Designed web scrawlers via Scrapy and Scrapy-Redis for the massive ID image dataset collecting work. Conducted image data ETL pipeline via OpenCV and Pandas. Constructed a fast-RCNN via Tensorflow on the image sets for model training and achieved 80% accuracy.
The project is an order management system and it's built by Maven, Springboot. In this project, customer can Create/Read/Update/Delete/Pay for orders, retailer can Read/Update order, Create/Read/Update/Delete product category/information. retailer can get notification when new order comes in, token authentication are enabled for validate users. I Designed redis cache to improve read/write efficiency.
WebdesignYelp Camp like website, it's created by node.js and mongodb. Basic crud functions.
WebdesignUtilize d3 to build a simple network graph for data visualization.
Data visualizationIn this work, we aim to build a system that can recommend musical hooks itself based on the following information: 1) latent representations of the hooks, 2) the music metadata related to the hooks (e.g. rhymes, repetitive part of the lyrics, genre, instrument, artist, etc), and 3) interaction between users and music. In order to incorporate external information such as the abovementioned 2) and 3), we introduce knowledge graphs into the system. In doing so, the performance of the system is expected to be augmented and the recommendations of the system can be explainable to the users.
KnowLedge Graph