Zexuan Zhou

ac2inc@r.postjobfree.com 765-***-**** LinkedIn: in/zexuan-zhou-5248b184/

OBJECTIVES: Seeking 2018 Fulltime Position in Data Science/Machine Learning/Artificial Intelligence SKILL SUMMARY

● Statistics:

Probability/information theory, Statistical inference Stochastic process, Bayesian inference, Time series

● Machine Learning: Regression: Ridge regression, Lasso, Kernel regression, Feature selection Classification/Clustering: Logistic regression, Decision trees, kNN Optimization: Gradient descent(SGD, Batch, Adam, Momentum, Coordinate descent) Deep Learning: DNN, CNN, RNN, LSTM

● Programming: Python, R, SQL, Spark, Scala, Java, Hadoop, Git, Linux EDUCATION

University of Washington, Department of Statistics, Seattle, WA 09/2016 - 06/2018 M.S. in Advanced Methods and Data Analysis.

● Relevant Coursework (In Progress): Applied Regression, Design and Analysis of Experiments, Foundations of Machine Learning, Statistical Inference, Statistical Computing, Intro to Artificial Intelligence Purdue University, College of Science, West Lafayette, IN 08/2013 - 05/2016 B.S in Mathematics and Statistics. GPA: 3.96 (Highest Distinction) EXPERIENCE

Intern, Meituan, Beijing, China 06/2017 – 09/2017

Group: Recommendation Algorithm R&D under Hotel & Tourism Business Group

• Implemented item based collaborative filtering with MapReduce for time-varying item candidate generation. Developed recommendation strategy utilizing geospatial information (geohash) and user profile information

(users’ preferred item category). Designed/performed online A/B testing and successfully launched an improvement with 2% CTR increase.

• Participated in DNN experiment of item candidate generation for recommendation. Mainly in charge of feature engineering. Developed data pipelines for PB-scale user history data with Hive/SQL. Optimized the efficiency with Spark transformations using Spark RDDs, Spark SQL and Scala. Project, JData Algorithm Competition - Prediction of customer purchase intention 05/2017 – 06/2017 Algorithm competition held by Chinese e-commerce group Jingdong (JD)

● Built classification models using XgBoost, logistic regression, SVM to perform prediction for customer purchase intention.. Designed sliding time window mechanism for feature engineering. Project, Pacman (Artificial Intelligence) 03/2017 – 06/2017 This is a course project for CSE 473: Artificial Intelligence

● Implemented algorithms such as BFS, DFS, A*, Minimax, Markov Decision Process, Reinforcement learning, HMM to provide policy for pacman to survive from the game with ghost as competitor. Project, House Prices: Advanced Regression Techniques (Kaggle Competition) 02/2017 Using residential information of 2390 houses as training set, predict housing sale price with 79 features

● Implemented and compared machine learning models such as Lasso, SVM, ridge regression. Random forest.

● Achieved R-square of 85.9%. Tools used: Python (sklearn, pandas, numpy, matplotlib) Project, Performance among Python Interpreters using CBD, University of Washington 12/2016 Designed complete block experiment and conducted contrast tests to evaluate performances of four different Python interpreters on three operating systems (R). This is a project for STAT 502: Experimental Design.

● Adopted additive model and interaction model to investigating treatment and block effects.