Lastest Update: Aug 2025   

Internship Experience

• Frost & Sullivan, Growth Consulting Department, Assistant Analyst(March 2025 - August 2025)
  Outline:
  • Participated in strategic research projects for the smart IoT industry, contributing substantially to three core projects: the obalt-nickel new energy materials industry analysis, high-performance hybrid chip market research, and smart IoT enterprise strategy diagnosis.
  • Engaged in the end-to-end process from industry status analysis and competitive landscape analysis to strategic path development.
  • Completed core tasks, such as building indicator analysis models, automating Wind data cleaning and processing using Python, and diagnosing enterprise issues and formulating strategies.
  • Collaborated with the team to align with client demands, assisted in the completion of multiple client presentations, and secured approval from company executives, ensuring effective alignment of research findings with business objectives.

• China Securities, Financial Planner Assistant(July 2024 - August 2024)
  Outline:
  • Industry Research: Led in-depth analysis of over 10 annual reports and prospectuses of leading companies in the industry, establishing a three-dimensional research framework for business, industry, and operations.
  • Model Application: Utilized DuPont Analysis and Porter’s Five Forces Model to quantitatively assess factors driving a company’s return on equity (ROE), the industry’s competitive landscape, and policy impacts.
  • Financial Diagnosis: Assessed the company’s profitability and cost control through horizontal comparisons of core indicators such as debt-to-asset ratios and gross profit margins, and analyzed industry development trends.

Research Projects

• Research on Marketization Allocation of Public Data Assets(Jan 2023-present)
  National Social Science Foundation Project
  Supervisor: Prof. Yong Tang
  • Outline: Deeply mining the characteristics of public and market-related data and further evaluate its inner value;
  • Responsibilities:
  Researching the impact of market sentiment on the stock market and quantitative analysis of market sentiment using deep learning and other AI models;
  • Achievement:
  Published LEET: stock market forecast with long-term emotional change enhanced temporal model in PeerJ Computer Science, 2024.( JCR Q1, IF 3.5) as First author and corresponding author.

• JP Morgan PTA Project,Quantitative Investment Strategy Development(Sep 2024 - Nov 2024)
  Outline:
  • Constructed a US tech stock dataset (2019-2024), with 10 years of historical data accessed through the yfinance API. Calculated logarithmic returns, performed volatility modeling, and conducted an asset correlation matrix analysis.
  
  •Implemented Markowitz mean-variance optimization and Black-Litterman model dynamic weight allocation, combined with CVXOpt to solve convex optimization problems.
  • Designed a Monte Carlo simulation system to generate 10,000 random portfolios to verify the robustness of the strategy.
  • Constructed a three-factor enhancement model: integrating market equilibrium returns, historical volatility, and momentum factors, achieving risk parity optimization through covariance matrix constraints.
  • Established a multi-dimensional evaluation system and completed a comparison with 5,000 random portfolios for verification.

• Imperial Vision Technology/Power System and Equipment Industry Research Institute(Sep 2024 - present)
  mmPowerHAR: A Framework Using mmRadar for Human Activity Recognition in Power Station
  Research Assistant (supervised by Prof. Jiang Hao, and Prof. Chen Zhenghua, Nanyang Technological University)
  • Outline:
  Based on the situations at the power distribution station, using mmWave radar evaluation boards to realise the determination of personnel posture and trajectory tracking;
  • Responsibilities:
  Designed a dynamic rotational position encoding mechanism to model the relative spatial relationships of limb motion trajectories using vector directions. .
  • Introduced a dual-view fusion strategy to integrate depth-azimuth (XOZ) and depth-elevation (YOZ) projections to construct 3D motion trajectories. .
  • Integrated a probabilistic sparse attention mechanism to reduce computational complexity from O(L²) to O(L ln L).
  • Modeled relative spatial relationships of limb motion trajectories using dynamic rotational position vectors.
  • The system has been verified in a real substation, providing a practical solution for power industry safety monitoring.
  • Achievement:
  Authored a research paper and submitted to IEEE Transactions on Power Delivery.

• Stock Price Prediction Research Based on RoBERTa-Informer(Jun 2023-Jun 2024)
  Supervisor: Prof. Yong Tang
  • Responsibilities:
  Project proposal and opening report writing, algorithm construction and experimentation, completion of project conclusion report and defense PPT production;
  • Achievement:
  Received provincial scientific innovation fund of ¥10,000for research, excellent conclusion(Only one in Economic and Managememt college).

• Quantitative Finance and Fintech under Artificial Intelligence (Application of Machine Learning in Business Data Analysis and Stock Market Prediction)(Aug 2023-Sep 2023)
  Provincial Undergraduate Innovation and Entrepreneurship Training Programme
  Supervisor: Prof. Patrick Rebeschini, Oxford University
  • Responsibilities:
  Exploring the application of deep learning models based on self-attention mechanism in quantitative trading

Under Construction