Machine Learning Data Analytics

Zhonghao (Charles) Wang

347-***-**** ******@********.*** New York, NY https://www.linkedin.com/in/zhonghaowang/ EDUCATION

Columbia University New York, NY

Master of Science in Operations Research GPA 3.83 (rank top 5 out of 197) Dec 2023 Relevant Courses: Asset Allocation, Algorithmic Trading, Business Analytics, Data Analytics (Python), Applied Analytics (SQL), Probability and Statistics, Stochastic Models, Machine Learning, Optimization, Simulation Shanghai Jiao Tong University (Top 3 University in China) Shanghai, CN Bachelor of Economics in Finance GPA 3.88 & Bachelor of Engineering in Industrial Engineering GPA 3.72 Jun 2021 SKILLS

Programming Languages: Python (Packages: Pandas, SciPy, NumPy, Matplotlib, TensorFlow, Pytorch, NLTK, Flask, Jupyter Notebook, Streamlit, Seaborn, ggplot2, CPLEX, Gurobi…), SQL, R, C++, C#, JavaScript, Java Software & Plarform: Excel, AWS, BigQuery, Looker Studio, Tableau, Matlab, Power BI, Apache Spark, Databricks Analytics Tools: Experiment Design (A/B Testing, Multivariate Testing), Feature Engineering, Dimensionality Reduction ML Models: Neural Networks (RNN, CNN, LSTM, GRU), Transformers (BERT, GenAI, Pathways), Random Forest, Gradient Boosting (XGBoost, LightGBM), SVM, GANs, PCA, ARIMA, Calibration, Multi-agent Systems, HMM, Bayesian Networks WORK EXPERIENCE

Pinya LLC New York, NY

Data Scientist Mar 2024 - Present

● Design and optimized predictive models, including Gradient Boosting and Random Forest to identify high-demand periods and time-series model, including ARIMA, LSTM and Prophet, to forecast monthly sales with 92% accuracy in Q3 2024

● Leveraged Apache Spark for distributed data processing, efficiently managing over 10 million rows of sales data monthly

● Created interactive dashboards with Python to visualize forecasted trends alongside historical sales for actionable insights Intrua Financial New York, NY

Data Science Intern Jun 2023 - Aug 2023

● Led the development of an asset allocation model based on linear programming using Gurobi to identify the mathematically optimal portfolio weight combination with improvement of 5% monthly return in the back testing during 5-year historical data

● Analyzed the model’s robustness and adaptability to various thematic tilts simulating the ESG investments using grid search

● Collaborated with the sales team to leverage data visualization platform to showcase the performance and robustness of model Dorfman Value Investment Shanghai, CN

Statistical Analyst Jun 2021 - Jun 2022

● Developed an automated trading framework for the bond market by leveraging reinforcement learning techniques, including Q-learning and policy gradient methods, to model decision-making under uncertainty

● Enhanced the trading framework by transforming the trading problem into a stochastic control problem by applying Bellman’s principle of optimality which reduced the model calculation time by over 10% on average in Bloomberg Terminal data

● Fine-tuned Large Language Models (LLMs) using PyTorch frameworks to analyze unstructured text data, enabling the trading framework to track updated news, generate automated insights, and enhance decision-making processes RESEARCH EXPERIENCE

Ad Recommendation System Research Based on Neural Collaborative Filtering New York, NY Columbia University Feb 2023 - May 2023

● Preprocessed datasets using NLP techniques (e.g., Tokenization, Bag of Words, TF-IDF, N-Grams) for feature engineering

● Designed and implemented a Neural Collaborative Filtering (NCF) network combining matrix factorization and neural networks and achieved a 20% improvement in hit rate on Amazon and Alibaba implicit feedback datasets Unveiling the Interplay Between On-demand Rides and Weather Patterns in NYC New York, NY Columbia University Oct 2022 - Dec 2022

● Integrated past 5 years of on-demand rides data and weather data utilizing SQL and SQLAlchemy to establish Object-oriented databases and optimize query efficiency by 30% and enabling faster retrieval for demand forecasting

● Established a versatile interactive interface, featuring seven distinct diagram groups and an integrated dashboard using Looker, that effectively conveyed the dynamical relationships of trips and weather, presenting a holistic view of data

● Forecasted high-paid ride start areas with accuracy of 97% and increased Uber drivers’ income by 30% in simulation SSE 50ETF Option Trading Strategy Based on Genetic Algorithm Shanghai, CN Shanghai Jiao Tong University Mar 2021 - Jun 2021

● Utilized the GARCH model to predict stock variance and applied the Black-Scholes equation to calculate option prices

● Developed an option trading strategy using genetic algorithm and achieved Sharpe Ratio of 1.2 and max drawdown of 20%

Contact this candidate