Hello World!

Oops, photo not loaded.

Qi Song
E-mail:hfjohn123@live.com
Tel: +1-(949)414-7432

Full time intern at ChenMed's Analytical Development team. Master of Data Science student of UCI. Graduate from Rutgers University in 2020 with a BS Computer Science degree. Actively looking for a full-time job.

This Website!


This website was made by myself from scratches. (Of course, I did use frameworks and libraries.)
Here is the list of frameworks and libraries I used:
  • Springboot+Thymeleaf
  • Jquery
  • BootStrap
  • Typed.js
  • Pace.js
  • Fullpage.js
  • Animated.js
  • Highcharts

And here is the link to the Github Repository. Please use the source code according to GPL licence.

Kinship Classification Ablation Study

This project was sponsored by Indago LLC, a company founded by former FBI agents.
The data set we used for this project was FIW.
In this project, we applied transfer learning on the dataset with paired images of portraits. We analyzed how the imbalanced dataset would affect the result. The resulting models can accurately classify the kinship between two blood-related persons.
You can click on the image for the report. And there is a link to the dataset and IPython notebook.

Rurl—— Short Url Generator

This website is a short URL generator that transforms your long URL into a short one so you can share it with others more easily.
As an example, this Demo collects UA (user agent) info from the request. This is an example of how could a website like this benefits the business.
MVC and HTTP Served by SpringBoot. Spring Data as ORM and MongoDB for data Storage. Design and function based on Restful API.
Click the image to get to the Demo. you can achieve the source code through the navbar of the demo site.

VGG-Like Neural Network Performance Research

Saying research but this was actually very shallow since this was my first DL project.
The data set we used for this project was here.
Basically we built a solid VGG-like model with Tensorflow, Trained and validated it with different training sets and test sets to see its performance and in the end concluded and analyzed the reason for the difference among the result, meaning what cause the difference of the accuracy.
You can click on the image for the IPython NoteBook, report and presentation files.

And More?

Please check my GitHub! I have many of my course works posted there.