Greater Cambridge Area
• Reviewed academic literature on machine learning methods (Autoencoders, Word2Vec, Seq2Seq, Transformer etc.) • Proposed a Seq2Seq-based LSTM approach: treated stock feature embeddings analogously to word vectors to forecast returns • Implemented and trained the model in Python (using PyTorch and scikit-learn) on 100 stocks, hourly data over 2 years • Generated predictive outputs, studied performance limitations, and identified potential improvements • Built competence in quantitative finance modelling, time-series data handling, ML algorithm adaptation, and research communication
• Analysed potential investment projects and products by investigating market backgrounds, trends, and competitive landscapes, evaluating positive outcomes, limitations, and risks to inform strategic decision-making • Provided data-driven suggestions on investment viability, leveraging economic analysis to support company growth initiatives • Investigated the entrepreneurial landscape in China and authored the report 2025 Report on Chinese Entrepreneurs
• Studied basic rationales about machine learning, and the working processes of several kinds of neural networks. • Learned about basic PyTorch codes, tested and developed projects using PyTorch • Researched publications on PINN, and participated in group study about relevant implementation methods to simulate the flow process of air in a closed unknown space given certain boundary conditions.
• Taught mathematics and conducted mock interviews to two classes of 20 to prepare for Oxbridge’s Maths assessments • Participated in several lectures to introduce Cambridge’s application process, mathematics course and university life • Improved the average score of TMUA of the class by 2 (out of 9)
• Collaborated with instructors to design and enhance teaching plans and materials, tailoring content to meet student needs • Managed a classroom of 20 students, providing individual attention and addressing student inquiries • Assisted in grading assignments and delivering constructive feedback, fostering students' academic comprehension