I’m a Ph.D candidate in Tsinghua-berkeley Shenzhen Institute, Tsinghua University, with research field on machine learning, data science, finance, economics, statistics, natural language processing, and correspondingly interdisciplinary research. If interested, please feel free to email me at keqinoly@163.com for further detailed information.

I have achieved my master’s degree of Management in June, 2023 from China Institute for Studies in Energy Policy, affiliated to School of Management in Xiamen University, with a major in Technology Economics and Management. My research area during Master period is Artificial Intelligence and Energy Finance, and dissertation topic is “Forecasting crude oil futures prices based on text-mining and machine learning techniques”, supervised by Xu Gong (龚旭). Previously, I have obtained a bachelor’s degree of Management from Business School of Central South University in 2020, with a major in Information Management and Information System, supervised by Jiangqiang Wang (王坚强).

My research interest includes Operation Research, Reinforcement Learning, Energy Management and Policy, Energy Finance, Climate risk, and Oil Prices Prediction. I have published 10 papers at the leading journal in finance and operation research.

📖 Educations

• 2023.09 - present, Ph.D, Institute of Data and Information, Tsinghua Shenzhen International Graduate School, Tsinghua University, Shenzhen, Guangdong, China.
• 2020.09 - 2023.06, Master, School of Management, Xiamen University, Xiamen, Fujian, China.
• 2016.09 - 2020.06, Bachelor, Business School, Central South University, Changsha, Hunan, China.

🔥 News

• 2023.10: My collaborative paper “Exploring City-Level Scholar Funding Service Impact: A Data-Driven Decision-Making System based on Difference-in-Differences Method” has been accepted by 2024 IEEE International Conference on Service Operations and Logistics, and Informatics!
• 2023.10: My collaborative paper “Conservative Periodic Reallocation Framework: Deciding research resources for research projects based on network DEA” has been accepted by Conference of 2024 ICCET!
• 2023.10: My Paper A new hybrid deep learning model for monthly oil prices forecasting is accepted by Energy Economics, a leading journal in Energy Finance with ABS-3 star!!
• 2023.09: 🎉🎉 I finally become a student of Tsinghua University, Congradulations!
• 2023.05: My another collaborative paper “Climate change attention and carbon futures return prediction” has been accepted by Journal of Futures Markets!
• 2023.02: My collaborative paper Carbon Information Disclosure and Bond Credit Spreads is published on Journal of Management Science (Chinese)!
• 2022.10: Just receive the Ph.D offer from Tsinghua University!
• 2022.09: Just take the IELTs examination!
• 2022.08: I participate in two high-level Summer Schools held by School of Economics and Management, University of Chinese Academy of Sciences.
• 2022.07: My Paper The role of textual analysis in oil futures price forecasting based on machine learning approach is published on Journal of Futures Markets, a leading journal in Finance with ABS-3 star!
• 2022.04: I join a famous Open-Source organization in China, DateWhale, congratulations!
• 2021.12: I win the China National Scholarship and Liu YuBin Graduate scholarship.
• 2021.08: My Paper What drives oil prices?—A Markov switching VAR approach is published on Resouces Policy, Web of Science with ABS-2 star!
• 2020.09: 🎉🎉 I get enrolled in Xiamen University, a prestigious university in China.

📝 SCIENTIFIC RESEARCH

📖 Published Papers

Information Processing \& Management (2025)

TEDRec: Transformer-based scientific collaborator recommendation via textual-edge dynamic network modeling.

Guan, K., Huang, W., Chen, T., Chan, WKV.* (2025).

• This paper is recently accepted by Information Processing \& Management (CCF B, ABS-2 star).
• We implement a novel academic collaborator recommendation framework called TEDRec, comprehensively modeling the collaboration network from three key elements, i.e., temporality, textual edges, structural relationships.
• Our data construction method can provide an innovative idea to build high-quality datasets for other fields.
• Conclusion: The proposed framework does achieve superb performance over baseline models across all evaluation metrics, indicating excellent generalization and robustness.

Energy Economics (2023)

A new hybrid deep learning model for monthly oil prices forecasting.

Guan, K., Gong, X. * (2023).

• This paper is recently accepted by Energy Economics (ABS-3 star).
• We introduce an independent module to preprocess IMFs generated by VMD to prevent the over-fitting problem and thus improve the forecasting precision.
• Our data augmentation can provide an innovative idea for other fields challenging to augment data artificially.
• Conclusion: The preprocessing of decomposed series is beneficial to capture temporal general feature patterns, thereby helping to produce more accurate and robust forecasting results than the competing benchmark models.

Journal of Futures Markets, 2022

The role of textual analysis in oil futures price forecasting based on machine learning approach
Gong, X., Guan, K.*, Chen, Q. (2022).

Project

• This work is published on Journal of Futures Markets (ABS-3 star).
• We creatively introduce text mining techniques (LDA, CNN, FinBERT, TextBlob) to forecast oil futures price.
• We offer an innovative approach to capture the trend of oil futures prices based on natural language processing and machine learning.
• Conclusion: The textual features and financial features are complementary in improving forecasting performance, and the improvement can be maximized by incorporating all features.

Resouces Policy, 2021

What drives oil prices?—A Markov switching VAR approach.
Gong, X., Guan, K., Chen, L., Liu, T. *, Fu, C. (2021)

Project

• This work is published on Resources Policy (ABS 2-star)
• We construct a five-variable Markov switching vector auto-regressions model based on five driving factors.
• We build this model to study the impact of different oil shocks on oil prices and analyze the factors under different regime conditions.
• Conclusion: The oil inventory and speculative demand have more significant effects on the oil price fluctuation than the oil aggregate supply and demand, and multiple factors have cumulative effects on the oil price fluctuation and the their intensity change under different regimes.

📖 Working Papers

Remote Sensing of Environment (SCIE-Q1)

Classification of small sample vegetation based on hyperspectral remote sensing and transfer learning
Huang, W., Guan, K.*, Hu, Y.

• This working paper is under review at Remote Sensing of Environment (SCIE-Q1).
• We provide a framework to efficiently perform hyperspectral classification of small samples of vegetation, which reduces model overfitting and diminishes category imbalance interference.
• We adopt transfer learning to alleviate spectral data shortage and innovatively consider cross-domain learning of spectra.

📖 Recent work

💬 Academic Conference

• 2022.08, International Conference on Climate and Energy Finance in 2022. (Oral Presentation: A new hybrid deep learning model for oil prices forecasting).
• 2020.10, The 11th China Energy Economic and Management Academic Annual Conference in 2020. (Oral Presentation: What drives oil prices?—A Markov switching VAR approach).

🏫 Summer School

• 2022.07, Asian Summer School in Econometrics and Statistics, University of Chinese Acadamy of Sciences. [Certificate]
• 2022.07, The Third Applied Economics Summer School, University of Chinese Academy of Sciences. [Certificate]

⛳ PROJECTS

🔍 Professional skills

• Achieve ML, DL and NLP specialization certification by Stanford University through Coursera. [ML][NLP][DL]
• Finish AI expert creation camp in PaddlePaddle Pilot Group of Baidu, and award Excellent learner and Excellent leader. [Certificate 1], [Certificate 2], [Note]
• Achieve the “News recommodation system practice” via DataWhale organization and award Excellent leader. [Certificate]
• Achieve the “NLP competition practice (E-commerce search)” via DataWhale organization and award Excellent learner and Excellent leader. [Certificate 1], [Certificate 2], [Note]
• Achieve the “Data visualization Matplotlib” via DataWhale organization and award Excellent learner and Excellent leader. [Certificate 1], [Certificate 2], [Note]

💡 Open-Source Projects

• whale-quant: A quantitative open source course, which can help people quickly master quantitative financial knowledge and the ability to use Python for quantitative development, with DataWhale open source organization. [Link]
• Undingable-Optimization: A optimization methods tutorial for elucidating the theory and algorithm implementation of operation research and optimization algorithms with DataWhale open source organization. [Link]
• Python Machine Learning: A briefly practical course for implementing simple machine learning in Python.
• Write a tutorial 100 questions take you to get started with the basic opretions of Python language on the HeyWhale platform, and obtain 4.9k+ reads and 170+ forks.
• Write learning notes about Eat pytroch in 20 days on the HeyWhale platform, and obtain at least 1k+ reads. [Note 1], [Note 2]

🪧 National-level Innovation and Entrepreneurship

• 2018, A creative cultural product design and sales platform relying on Hunan embroidery elements, funded by Ministry of Education in China with 13000 yuan.
• 2019, Online product review recommendation mechanism based on data mining technology, funded by Ministry of Education in China with 8000 yuan.

💻 COMPETITION EXPERIENCE

🎙 Mathematical modeling

• 2022, The Third Prize in MathorCup Mathematical Contest in Modeling - Big Data Competition. [Certificate]
• 2021, The Honorable Mention in ShuWei Cup IMCM (Faculty Advisor). [Certificate]
• 2021, The Second Prize in Shuwei Cup Mathematical Contest In Modeling (MCM). [Certificate]
• 2020, National Third Prize in “Huawei Cup” The 17th China Post-Graduate MCM. [Certificate]
• 2019, Meritorious Winner of Interdisciplinary Contest In Modeling (Top 8% in global competitors). [Certificate]
• 2018, National Second Prize in Contemporary Undergraduate Mathematical Contest In Modeling (Round top 4%).

🖥️ Data Mining Competition

• 2022, Alibaba Tianchi E-commerce Search Algorithm Competition. [Certificate]. (NLP, 128/2771)
• 2022, Global AI Commodity Title Entity Recognition. (NLP+NER, 158/1700)
• 2022, Cattle image instance segmentation. [Certificate]. (CV, 21/790)
• 2022, Real scene tamper image detection challenge in TianChi. [Certificate]. (CV, 167/1149)
• 2022, Xiamen International Bank Financial Innovation Competition. (RSs, 63/1529)
• 2021, WSDM (Top Conference) IQiyi User Retention Forecast. (RSs, 65/991)

Abbreviation: NLP, Natural Language Processing; NER, Named Entity Recognition; CV, Computer Vision; RSs, Recommendation System.

📱 Innovation and Entrepreneurship Competition

• 2022, The Bronze Award in The 8th Fujian “Internet+” College Students Innovation and Entrepreneurship Competition. [Notice]
• 2021, The Silver Award in The 7th Fujian “Internet+” College Students Innovation and Entrepreneurship Competition. [Certificate]

🎖 HONORS AND AWARDS

🏅 Scholarship

• 2024, TBSI Leaders of Tomorrow Scholarship (Merit Scholarship), Tsinghua University.
• 2023, Anta Graduate Scholarship, Xiamen University. [Certificate]
• 2022, Hong Xin Graduate Scholarship, Xiamen University. [Certificate]
• 2021, China National Scholarship (Top 0.2% in Chinese students). [Certificate]
• 2021, Liu YuBin Graduate Scholarship, Xiamen University.
• 2019, The Third Prize Scholarship, Central South University. [Certificate]
• 2018, Jiang Weiying Scholarship, Central South University. [Certificate]
• 2018, The Third Prize Scholarship, Central South University. [Certificate]
• 2017, Century Haixiang Reward and Attendance Scholarship, Central South University. [Certificate]
• 2017, The Third Prize Scholarship, Central South University.

✨ Honors

⏰ Self-evaluation

• A dedicated, detailed and capable research-oriented student. Having extensive knowledge of the cutting edge of ML and NLP techniques and being able to apply it in academic research.