Education

University of Virginia, Charlottesville, VA

Master of Computer Science December 2019

Related Courses: Data Structure and Algorithms, Compiler, Software Engineering, Machine Learning, Natural Language Processing, Deep Learning for Visual Recognition, Statistical Learning and Graphical Models, Cloud Computing…

current GPA: 3.9 / 4.0

Nanjing University, China

Master of Electronics and Communication Engineering(postgraduate recommendation) July 2018

GPA: 4.38 / 5.0

Bachelor of Electronic Information Science and Technology July 2015

GPA: 4.26 / 5.0, Top 15%

Honors: First Prize of Postgraduate Scholarship(twice) 2015&2017

​ Second Prize of the People’s Scholarship2012

Experiences

Intern, July. 2017 - Nov. 2017

Basic Platform Data Mining Group, Yiducloud (Beijing) technology Ltd. (Startup)

Project 1: Medical Records Scoring

• Achieved AUC of 0.9 by applying GBDT model to learn artificial evaluation criteria
  
• Utilized Spark to accelerate the training speed
• Helped launch the product to dozens of IIIA hospitals for generating evaluation results automatically

Project 2: Template Learning

• Evaluated the weight of the nodes and its value(XPath) in the CDA documents by performing a similar approach as tf-idf

• Used the important XPath as the keys to assist colleagues in producing the semi-structured output

Projects

Hair DyeMarch. 2019 - May. 2019

• Responsible for the neural network implementation, part of data loader and deploying to Android Application
• Deployed the Pytorch model to Android Application by using ONNX and Fritz framework
• Achieved F1 Score of 0.9 and IoU score of 0.817 on Figaro 1K dataset by using HairMatteNet model
• Git link: https://github.com/aobo-y/hair-dye
• Git link for Android App: https://github.com/quq99/hair-dye-android

MeggyJava CompilerMarch. 2019 - May. 2019

• Used the JLex and JavaCup tools for generating the lexer and parser
• Implemented a semantic analyzer for Symbol Table and Type Checking
• Implemented AVR code generation visitor

SouthPark ChatbotOct. 2018 - Dec. 2018

• Designed and implemented the evaluation part
• Achieved BLEU score of 15 on average, 75 the best by using the persona-based Seq2Seq model
• Trained a chat robot with personality and can perform diverse responses by using Transfer Learning and Beam Search
• Git link: https://github.com/quq99/SouthPark-Chatbot

Opioid Addiction Crisis in VirginiaOct. 2018 - Dec. 2018

• Performed anomaly detection in high dimension data by using Isolation Forest algorithm
• Manipulated data visualization in high dimension data by applying K-Means algorithm
• Improved the learning speed of 15% by using Principal Component Analysis (PCA) to reduce the input dimension
• Git link: https://github.com/quq99/Opioid_Addiction_Crisis_in_Virginia
• Video link: https://www.youtube.com/watch?v=Cq-OQ9UcBRo&feature=youtu.be

Agent Learning in OpenAI gymNov. 2016 - Dec. 2016

• Selected three environments in OpenAI gym: Acrobot-v1, CartPole-v0, pong
• Implemented reinforcement learning algorithms including CEM, PG, DQN and tested them in OpenAI gym
• Considered feature engineering to preprocess the input data, improved the learning speed twice
• Improved score by 5% by applying the actor-critic method to PG algorithm,
  

Skills

C++, Java, Python/ Tensorflow, Pytorch, Scikit-learn/ Git , shell script