Course Work Project (Texas A&M University)

1. Deep Learning, Computer Science & Engineering (636), 2020 Fall

• Constructed SeNet92 and DPN92 to classify Cifar10 Dataset. 

2. Machine Learning, Computer Science & Engineering (633), 2020 Fall

• Led team of 5 on designing and disseminating CNN for COVID-19 Diagnosis. 

3. Stochastic Dynamic Programming, Industrial & System Engineering (637), 2020 Spring 

• Led team of 3 on constructing IRL with Zero-Order algorithm for the bus company. 

4. Analysis of Algorithms, Computer Science & Engineering (629), 2019 Fall 

• Constructed and compared Dijkstra’s algorithm and Kruskal’s algorithm. 

5. Numerical Methods in Partial Differential Equations, Mathematics (610), 2019 Spring

• Constructed numerical algorithms with finite difference and finite element methods for solving partial differential equations.

6. Deep Learning Application, Mathematics (689) Department, 2018 Fall

• Constructed a CNN-based model to diagnose the fault of fused deposition and compared it to the SVM method

Research Experience

1. Multi-agents Game with Imperfect Information (IRL, Game Theory), Jan 2021 - Aug 2023

• Achieved better strategy under different business cycles with the POSG model.

• Increased the five burger companies’ profits by 7%.

• Presented at the 2022 INFORMS Annual Meeting.

• Publication: 2nd author, under review in Management Science, Survive and Thrive: Strategic Firm Expansion under Partially Observed Business Cycles.

2. A Classification for One-agent Hidden Brands (Human-Autonomy Interaction, Big Data), Jan 2022 - Present

• Supervised undergraduate students’ capstone project.

• Distinguished human drivers from safe and unsafe driving through the Hidden Semi-MDP model.

• Examined extensive baseball data within Major League Baseball (MLB).

• Publication: work in progress, NSF Project with Human Factors and Machine Learning Lab.

3. One-agent Game with Imperfect Information (IRL), Sep 2020 - Dec 2020

• Out-performed ’scipy.optimize’ in global optimum by polishing the soft policy gradient algorithm.

• Achieved accurate estimation of the reward parameter and proved the algorithm's convergence.

• Reduced the price by $20+ per month for one bus in engine replacement cost by POMDP model.

• Publication: 4th author, IEEE Transactions on Automatic Control, Structural Estimation of Partially Observable Markov Decision Processes.

4. Staffing Management for Self-scheduling Service, Sep 2019 - Dec 2019

• Proposed a uniform-price auction scheme for staffing in self-scheduling businesses.

• Achieved lower costs in short-term staffing compared with the fixed-wage model.

• Publication: 2nd author, the Best Paper in the 2021 IISE Transactions Focus Issue on Supply Chain and Logistics, Uniform-Price Auctions in Staffing for Self-Scheduling Service.

SKILLS

1. Languages: English and native proficiency in Mandarin (Chinese)

2. Programming: Python, Matlab, PostgreSQL, C, R, LaTeX, Google Cloud

3. Tools: POMDP, IRL, MDP, Data Analytics, Problem-Solving, Algorithms, ML, DL, Game Theory, Pandas, NumPy, SciPy, PyTorch, Ray, Matplotlib, TensorFlow