Hello, I'm Xiaoxiang Ma

I am a M.Eng. student at Cornell University studying operations research and expecting to graduate in December 2021. I have experiences working with data science & machine learning in fields including telehealth, finance, and aid allocation.

6072299675
xm53@cornell.edu

Latest Work

A collection of internship / personal project / volunteer work

Symptom Checker MVP design

Designed and implemented the reward system and Markov Decision Processes of a symptom checker chatbot powered by Noisy DQN (Reinforcement Leaning).

  • Duration: 8 Weeks (2020 Summer)
  • Service: AI development
  • Type: Remote Internship
Reinforcement learning Chatbot design DQN Noisy DQN Python Markov decision process Reward design Keras Jupyter notebook

Identifying Vulnerable Refugees

Conducted statistical analysis of refugee chatbot data and created random forest and regression models to improve vulnerability classification.

  • Duration: 6 months
  • Service: Data Analysis
  • Type: Pro Bono
Data mining Data cleaning Chabot design Data infrastructure Python Statistics Logistic regression Random forest Jupyter notebook

Predicting Startup Longevity

An analysis of key factors to longevity in the U.S. Startup industry using Logistic Regression.

  • Duration: 5 Weeks (2020 Summer)
  • Service: Market Research / Analysis through data
  • Type: Independent Project
Experiment design Python Market research Variable selection Statistics Logistic regression ROC Webscraping Jupyter notebook

Strengths in Badminton Using BWF data

A deep dive into service, point lead, and consectutive points and how they affect the strategic/mental game world-class players.

  • Duration: 5 Weeks (2019 Spring)
  • Service: Sports Analytics
  • Type: Independent Project
Data visualization Python Logistic regression Data pre-processing Statistics Webscraping Jupyter notebook

Classifying Low-credit Loanees

Designed loanee scorecard model for China Foreign Economy and Trade Trust Co., Ltd. using an aggregate of loan data, demographic data, and property data.

  • Duration: 6 Weeks (2019 Summer)
  • Service: Model design, training, and testing
  • Type: Internship
Scorecard model Logistic regression AUC ROC Statistics Variable selection SAS SAS Enterprise Guide SAS Enterprise Miner SQL Python