WELCOME! This website serves as my personal portfolio. It includes most of the projects I have completed, individually or as a group, during my academic and professional time.

"Projects" include most of the STEM-related projects I've completed. Topics included are: Data Science, Data Analytics, Machine Learning, programming languages, etc. Most of them were school-related projects.

"Others" include mostly non-STEM projects, presentations, and research paper replications (replicating results from academic articles).

"About Me" is where you can find more information about me. It includes other fields that I am interested in, as well as my professional experience.

STEM Projects

Data science & data analytics

  • Machine Learning: Best Frac Crew
  • This project was led by UT Austin Inventors Program and partnered with ConocoPhilips, a natural gas liquid company. We as a team of 5 designed an automated process for ConocoPhilips to determine their best hydraulic fracturing contractor(s). We were given over 360,000 rows of data of the contractors and their hydraulic fracturing equipments. The project went through 3 minimum viable products based on client's feedback and UT faculty's guidance. The final product used a K-means algorithm to provide an automated process of selecting the best contractor(s) based on selected features. And we visualized this product using a Tableau dashboard. You can find our team presentation to the UT students here.

  • Data Analytics: Airbnb Price Prediction
  • This is a class project from the UT Austin MSBA program. We analyzed a kaggle dataset of Airbnb listings in New York City to predict the price of a new Airbnb listing with certain features. In this project, our team performed a complete data analytics process, including EDA, data wrangling, model building, feature selection, etc. In the end we also ran an error analysis to see how we may improve the results.

  • Data Analytics: Hotel Booking
  • The goal was to analyze hotel booking information for two hotel franchises to better predict the hotel cancellation, which would be helpful for the hotel management system. Similar to the previous project, we performed EDA to the data and then ran a number of algorithms and models to train the data. We concluded the analysis with an evaluation of all models using certain metrics. The final presentation slides are here. And you can find a detailed report here.

  • Machine Learning: NBA Game Outcome Prediction
  • This is for the Advanced Machine Learning course. With a team of 4, we built a predictive model to make NBA game predictions, given two teams and their player rosters. The content is unique in a way that we used the players' historical statistics for predictions instead of just teams' information. In this blog post, you can read more about our project.

  • Optimization
  • Optimization was a required one-year course for the MSBA program. The first semeter was an introduction to optimization concepts and the Gurobi software, along with mostly linear optimization problems. The second semester involved neural network and non-linear optimizations. There are 6 projects included below:

    Linear Programming, Dedicated Portfolio We are building a dedicated portfolio that can hedge against interest rate risks and whose cash inflows match the outflows. (Problem statement; Report)
    Linear Programming, Integer Programming, Index Fund We are constructing an index fund to track the NASDAQ-100 index as close as possible. (Jupyter notebook)
    Quadratic Programming, LASSO This project tries to pick the best variables in a regression problem and compare the results with LASSO. (Jupyter notebook)
    Newsvendor, Stochastic Programming This project is an extension to the over-simplified "newsvendor" problem with more parameters. (Jupyter notebook)
    Airline Pricing, Dynamic Programming Given certain options for ticket classes and prices, we are building an overbooking policy for an airline company (how many overbook sales are allowed, etc.) so that it can maximize profits while keeping the overbooking costs at the minimum. (Jupyter notebook)
    Neural Network, Reinforcement Learning We are building an automated game player for "PopOut", a variation of "Connect-4". We used reinforcement learning to train the model so that the computer can learn from human moves and have a better chance of winning. (Jupyter notebook)
  • Unstructured Data Analysis: Music Streaming Platform
  • Unstructured data analysis includes mostly text and articles. In our project, we scrped 3,000 user comments of several music streaming apps from Apple Store (Spotify, Apple Music, etc.) and constructed a recommender for users with their preference parameters (contents, fees, etc.). Due to the limited time frame, this project was more of a prototype. But it still provided some interesting findings about the online review information. Concepts included here are Sentiment Analysis and Topic Modeling.

  • Social Media Network Analysis: Chinese Balloon Incident
  • Social media network analysis is a more specific field in unstructured data analysis. This project was about a time-sensitive subject. We scraped over 10,000 tweets online about the Chinese balloon incident happened in February, 2023. Then we preprocessed the data and performed several network analysis topics, including Centrality, Sentimen Analysis, and Topic Modeling. We were looking for the user(s) who may influence the direction of this incident the most, as well as what the public may feel and concern about the most from the incident.

  • Deep Learning: Music Generation using LSTM
  • The concepts we covered in the Deep Learning course (NLP, Neural Network, Adversarial Attacks) consist of a very tiny portion of the current Deep Learning community. The topics introduced in this project were also a learning process for us. Throughout this project, we followed along closely with multiple online articles and successfully reproduced the music-generation algorithms they discussed. We used Recurrent Neural Networks (RNNs) and Long Short-term Memory (LSTM) architecture to build a music generator. Training data were selected from kaggle. Read more on this blog post.

  • MSBA Capstone Project: Anomaly Detection and Time Series Analysis
  • The capstone project is a semester-long practicum course of the Business Analytics program. Our team of 5 partnered with Affinity Answers, a local data-driven marketing strategy company in Austin, TX. The object of this project is to build an anomaly detection algorithm to mitigate data handling errors and to recgonize user behavioral anomalies so that our client is about to produce a more accurate data-driven marketing solution to their customers.


  • SQL: COVID-19 mini project
  • This mini project tried to answer a few questions about COVID-19 using SQL and other Big Data tools available on Google Cloud Platform (GCP).

  • Processing (Java): Game Simulation
  • This is a class final project using Processing as a platform for game simulation. As described in this Github repo, our team of 3 were recreating the game "Space Invader" with a few modifications. Instead of displaying spaceships shooting against aliens, we designed a zombie-attacking game with a few power-up options. The final game demo includes music, sound effects, player manual, and game score history. You can visit the above repo, download the game and try it out!

  • Website Designing: Disease Prevention
  • This project was for an intro-level website designing course where we leanred HTML, CSS, JavaScript, PHP, and a few other web development languages. Due to the limitation of time, we were not exposed to a lot of practice of building a dynamic website. The project includes mostly static webpages with one PHP script (the questionnaire form in the "Contact Us" page). And due to the school restriction, the original domain names are no longer available. Please view the repo for each individual page instead.


Finance & Financial Analytics

  • Private Equity: YETI Leveraged-Buyout (LBO) Case Analysis
  • This case study is for the Private Equity course at UT Austin. Our team of 3 analyzed public information of YETI, Inc. We have conducted company, industry, risk, and peer performance analysis. Then we gave our investment thesis which might facilitate the company's growth. We have also quantified our analysis with a forecasting model, a DCF model and a comp analysis.

  • Applied Valuation: Estimating Beta via SCL and Comparable Analysis
  • Applied Valuation: Credit Risk and Ratings Prediction
  • The goal of this project is to determine what factors of a company may impact its credit ratings in the market. We have used WRDS data and some machine learning algorithms. Due to the time frame limitation of the course, our project was left incomplete. The report is a snapshot of our progress and noted the caveats of our models.

  • Asset Management: Index Replication
  • This project was to replicate the Dow Jones Industrial Average Index (DJIA) using WRDS data. Details can be found here.

  • Asset Management: Effect of Fund Size on Performance of Small-Cap Value Funds
  • Python script can be found here.

  • Asset Management: Portfolio Simulation
  • Python script can be found here.

  • Empirical Finance: R&D Replication
  • This was an individual project. In this project, I analyzed companies' financials from WRDS and created different portfolios based on their R&D expenses. These portfolios can be equal-weighted or value-weighted with other criteria. Then I calculated the returns and compared the CAPM alphas, Fama-French 3-factor alphas, and Sharpe Ratios between portfolios. The results are shown in this deck.


  • Directed Reading Program
  • The Directed Reading Program ("DRP") is an RTG program of the Mathematics department at UT Austin. I was priviledged to have my mentor Milad help with my research and presentation. My research for this program was about data analysis, with a focus on classification algorithms. My presentation was related to K-Nearest-Neighbors (KNN).

  • Information Management: Amtrak Data Management
  • The presentation was about data management for the public transportation organization, Amtrak, in a very high level. We analyzed Amtrak's data structure and used SQL to perform high level analysis.

Paper Replication

About Me

My name is Ziyu ("zeu-you") Wang and I'm orginally from Shijiazhuang, Hebei, China. My journey in the states began in 2015 in College Station, TX. In 2018 I embarked on my college journey at the University of Texas at Austin (Hook'em!). Five years later I graduated with BS Mathematics and MS Business Analytics, along with a Business Minor, a Computer Science Certificate, and a Computational Engineering and Sciences Certificate.

During my free time, I enjoy traveling and road trips, skiing, and watching shows and animes. I like to meet people from different backgrounds and countries.

Currently I am working as a quantitative management associate at a financial service company in North Carolina. And in the near future, I'm planning to work in the Asia-Pacific region. If you are interested in working in finance industry or have some working experience in the APAC region (or anything else), feel free to connect on LinkedIn!