CV
Curriculum Vitae
Zikai (Nathaniel) Wei
Quantitative AI researcher with a PhD from MMLab, CUHK. Industry experience at a top-tier hedge fund in Hong Kong and a top-tier asset management firm in mainland China. Research focus: end-to-end deep learning for active investing, adaptive factor models, and event-driven trading.
Education
Ph.D., Information Engineering — The Chinese University of Hong Kong, Aug 2018 – Jul 2023
MMLab; advisors: Prof. Xiaoou Tang, Prof. Dahua Lin
Thesis: Deep Learning for Predicting Real-World OutcomesM.Phil., Applied Mathematics — The Hong Kong Polytechnic University, Jan 2014 – May 2017
Advisors: Prof. Heung Wong, Prof. Cedric Ka-Fai YiuB.A., Finance and Banking — Shenzhen University, Sep 2009 – Jun 2013
Outstanding Graduate (rank 1/405); GPA 3.9/4.0
Experience
Quantitative AI Researcher — Present
End-to-end deep learning, adaptive factor modeling, market regime detection, and event-driven prediction. Working with Prof. Jian Guo on Janus-Q.Analyst, Investment — Top-tier hedge fund, Hong Kong SAR, Sep 2024 – Mar 2025
Multi-market monitoring, event-driven backtesting, global factor data integration (Bloomberg/Barra).Quantitative Researcher — Top-tier asset management firm, mainland China, May 2022 – Sep 2024
Deep learning strategies for CSI 300/800 and mid/small-cap indices; led quant AI platform development. Models include HireVAE and multi-agent GPT factor discovery.Research Intern — SenseTime Group Limited, Hong Kong SAR, Feb – Jul 2018
TCN/LSTM for financial futures; RL environment for limit order book data.Research Assistant — The Hong Kong Polytechnic University, Apr 2017 – Feb 2018
Deep learning jump detection across 11 global indices; regime-switching on high-frequency data.
Selected Publications
Selected work most relevant to quantitative AI and investment systems. Full bibliography: Publications · Google Scholar.
- Janus-Q: End-to-End Event-Driven Trading via Hierarchical-Gated Reward Modeling
- FinKario: Event-Enhanced Automated Construction of Financial Knowledge Graph
- E2EAI: End-to-End Deep Learning Framework for Active Investing
- HireVAE: An Online and Adaptive Factor Model Based on Hierarchical and Regime-Switch VAE
- Factor Investing with a Deep Multi-Factor Model
- Jump Detection in Financial Time Series using Machine Learning Algorithms
Skills
| Area | Details |
|---|---|
| Quant AI | End-to-end investment systems, factor models, regime detection, event-driven signals |
| Deep Learning | PyTorch; graph attention, VAE, reinforcement learning |
| Programming | Python, SQL, R, Matlab, C++ |
| Financial Data | Bloomberg, Barra, WIND, FactSet, Refinitiv |
Honors & Awards
- 2023 — China Merchants Group New Star Training Camp, Outstanding Individual
- 2019 — CUHK-SenseTime Joint Lab Excellent Contribution Award
- 2018–2023 — Postgraduate Studentship, CUHK
- 2015 — Teaching Postgraduate Studentship, PolyU
- 2013 — Outstanding Graduate, Shenzhen University (1st in Finance & Banking)
- 2012 — National Postgraduate Recommendation (examination waived)