Participated in the design of a macroeconomic indicator monitoring system, extracting data from databases using SQL (including daily stock and fund data, macroeconomic indicators, etc.), merging and transforming data into a presentable format using Pandas, and visualizing the obtained data using Dash.Contributed to the design of a technical analysis signal research framework, generating buy or sell signals, measuring the practicality of signals using metrics such as the Sharpe ratio, P&L (Profit and Loss), win rate, and maximum drawdown.Built a Convolutional Neural Network (CNN) model: Utilized MPL Finance to generate candlestick charts, constructed a neural network using PyTorch, supervised the training of candlestick charts, and predicted price movementsDeveloped a fintech-themed company scoring model: collected company data, including patents data, company descriptions, news, etc., and fine-tuned the Finbert Model for analysis and generated the score

Predicted actions of buy-ins or sales through deep reinforcement learning to maximize the profits given stocks’ historical dataUtilized python to carry out data preprocessing, environment setup, establish Deep Q-Network and adjust parametersImproved model performance around 10% on back testing dataset by neural network structure modification and hyperparameteroptimization

Worked as a product manager assistant at the “Technology Department” to help coordinate various major development projects