Welcome to Maggie's Page!
Huiying Mao (Maggie, 毛慧颖)
Data scientist / Statistician / Runner / Yogi
I am a data scientist at eBay focusing on online experimentation.
Before joining eBay, I were trained as a Postdoctoral Fellow at SAMSI.
I got my Ph.D. in Statistics from the Department of Statistics, Virginia Tech
Latest news:
I just completed Coursera course "Generative AI with Large Language Models." It illustrated the LLM basics, like transformer, and three stages of the LLM training pipeline: Base model --> Fine-tuning --> Reinforcement Learning with Human Feedbacks (RLHF). This course better prepared me in building LLM-based applications.[2023.10.10]
Our rencent work on Online Controlled Experiments "Clustering-Based Imputation for Dropout Buyers in Large-Scale Online Experimentation" is published at The New England Journal of Statistics in Data Science, please visit here for more details [2023.05.24]
My postdoctoral work on "spatial conformal prediction" is published at Journal of the American Statistical Association (JASA), please visit here or here for the details [2023.01.05]
In spatial applications, modeling non-stationarity and non-Gaussianity is hard. Making valid prediction a challenge. Here we do it without modeling, using conformal prediction techniques. Comments welcome!
Related Github repo / R package: https://github.com/mhuiying/scp
I joined eBay as a data scientist in the Experimentation Platform team [May 2020]
Thanks a million to Statistical and Applied Mathematical Sciences Institute (SAMSI) for training me as a postdoc. I learned a lot and became a researcher with more confidence for moving forward!
Just defended! Officially a Ph.D. now ✌ [2019.07.23]
Starting August, I will be a Postdoctoral Fellow at the Statistical and Applied Mathematical Sciences Institute (SAMSI), focusing on Games, Decisions, Risk and Reliability
Finished my periodic internship at Didi Chuxing, the largest ride-hailing company in China [2019.06.18]
While being there, I helped Didi to establish an accurate, robust, and explainable driver risk prediction system based on drivers' driving behavior profile.
The system can effectively reduce noise and extract information from the high-frequency mobile phone sensor data
For more details, please visit here.
Received a fruit bouquet from a team I mentored for their capstone project at the end of semester. Feeling so touched ❤️ Mentoring students is definitely one of my favorite things to do! [2018.12.08]
I won the most outstanding service award this year! I was awarded by Dr. Klaus Hinkelmann, our former department head. He also gave me his three volumes of Design and Analysis of Experiments as a gift! I feel so encouraged! 😊😊 [2018.10.30]
My department has seven awards to give on our annual Corporate Partner Conference. I won FIVE of them during my five years of Ph.D. life.
- Klaus Hinkelmann Award, Outstanding Service by Graduate Student, Department of Statistics, VT, 2018.
- Rose Costain Award, Outstanding Citizenship by Graduate Student, Department of Statistics, VT, 2017.
- Jess C. Arnold Award, Most Outstanding Teaching by Graduate Student, Department of Statistics, VT, 2016.
- Boyd Harshbarger Award, Most Outstanding First Year Graduate Student (Rank 1/33 in Ph.D. Qualify Exam), VT, 2015.
- Raymond H. Myers Award, Most Outstanding First Year Graduate Student in Linear Models and Design of Experiments (Rank 1/33 in Experimental Design), VT, 2015.
My 4th annual half marathon, check! I've been running half marathons each year since I start my graduate school. This year, I ran the Hokie half, held at Blacksburg, VA, the town of Virginia Tech. Let's go, HOKIES! [2018.09.23]
I just completed "Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization", taught by Andrew Ng, on Coursera. With plenty of intuitive illustration, Andrew gives hands-on instructions of state-of-art hyperparameter tuning strategy. This course is the second of the five-course specialization offered by deeplearning.ai. I've completed the 1st and 2nd, and I will continue learning the rest. [2017.09.28]
In the SAMSI summer workshop on Transportation Statistics held in Duke University, I've won the first-prize in student poster competition. Danni Lu, my academic sister, has won the second-prize. (Congrats on us :-)) [2017.08.17]
In this poster, I've presented the evaluation of high G-force event on crash risk using SHRP2 NDS data. For more information, please visit my research page.
I just finished my summer internship at Microsoft Knowledge and Growth, Risk team. It's a wonderful team to work with. Special thanks to my mentor, Yuting Jia, and manger, Justin Wang, who had offered me tons of guidance. [2017.08.04]
For this summer project, We improved the current fraud detection system's performance using entity profiling with fraud feedback. For more information, please visit my research page.