Mingyang Lou Email & Phone Number

San Francisco, CA, US

San Francisco, CA, US

New York, NY, US

Collaborate with JP Morgan to work on reinforcement learning on the optimal trading problem.

Berkeley, CA, US

• Conducted research on Diabetic Retinopathy screening with machine learning models
• Performed image preprocessing including rescaling, resize, augmentation, and enhancement
• Conducted K-Means and PCA to extract image features, trained models such as SVM, logistic regression, and LDA to predict DR level with 95% specificity and 73% accuracy
• Built a CNN model with multiple convolutional, max pooling, dense layers using Keras and TensorFlow

Berkeley, CA, US

• Data processing on 100,000+ news articles and S&P 500 index over 1781 trading days - Performed NLP (tokenize, remove stop words, word embedding, etc.) to extract information from financial news - Programmed in Python.
• Trained various machine learning models, such as the cross-validated random forest, neural networks (CNN, LSTM), and the boosting, to predict stock movements
• Applied the majority voting on previous model predictions. The final model outperformed the baseline with 58.26% accuracy

南京, 江苏, CN

• Built multi-factor models by constructing more than 200 industry-level factors on consensus estimate data(industry- and firm-level)
• Selected effective factors by simulating results and ranking indicators such as factor scores, information coefficients, returns and risk measure ratios
• Automated multiple steps in data processing and factor performance analysis process which significantly improved the efficiency of factor analysis process in the financial engineering team

• Improved the quantitative research system; researched, implemented and modified systematic trading strategies such as statistical arbitrage and factor-based strategies to expand quantitative content for JoinQuant’s.
• Built 4 trading interfaces that connect JoinQuant and prime brokers; wrote vectorized functions and modules to accelerate the process by 50%
• Wrote trading functions to improve the user experience

• Constructed a passively managed portfolio of cryptocurrencies by applying various regression methods to select components and features; researched and reduced the tracking error from 17% to 8%
• Assisted in the quantitative strategy research of hedge funds: conducted an event-driven strategy study on newly listed cryptocurrencies from different exchanges; performed statistical analysis on price series, and.
• Conducted independent research on the pricing of BTC/USDT options based on the Black-Scholes model in Python; assisted in CPPI strategy research
• Built a toolbox in Python, enhanced data processing and generated weekly performance reports automatically

London, GB

• Conducted audit testing and external confirmation on over 40 client’s bank accounts
• Performed financial statement reviews and integrated financial statement audits; identified risks associated with particular clients and designed tailored risk control procedures

Master Of Engineering - Meng, Ieor(Fintech)

Bachelor Of Science, Finance, Math

Summer School, Department Of Economics