Zeqing Qin

Zeqing Qin

HUST

About me

Hi! I am a postgraduate student in the school of Cyber Science and Engineering at Huazhong University of Science and Technology(HUST). My research interests are Machine Learning, Natural Language Processing, Security, and Multimodal Image Processing.

What’s News

  • Start working in HUST, Sep 1, 2022.
  • Start the internship in UTD, Apr 28, 2023.

Research Experience

Improve BERT-based LLM’s Performance on C/C++ Vulnerability Detection
Otc 2023 - Dec 2023
Supported by the National Natural Science Foundation of China
Project Leader and Algorithm Developer

  • I identified BERT-based models’ limitations in understanding C/C++ compared to interpreted language like PHP.
  • Designed an intermediary layer influenced by DNS for C/C++ and model integration.
  • Improved the performance of BERT-based model’s performance on C/C++ vulnerability detection, and the research paper is under review.

Image Registration Method Based on Evolutionary Computation
Aug 2020 – Mar 2022
Supported by the National Natural Science Foundation of China (Grant No.61772180)
Person in charge (Team leader and the only undergraduate student in the group)

  • Optimized the process of medical image registration with Evolutionary Computation.
  • Improved genetic algorithm by simulating the breeding process of Chinese three-line hybrid rice.
  • Achieved faster convergence speed and greater robustness. Published a paper at IEEE SMC2021.

Extracting Concept Prerequisite from Network Resources
Jul 2020 – June 2022
Supported by the National Natural Science Foundation of China (Grant No. 61977021)
Assistant Research (Assisted in the design and implementation of experiments)

  • Innovatively explored from the two dimensions of Wikipedia and the course introduction.
  • Dug out the primary dependencies between learning objects more accurately. Published a paper at SEKE2021.

Professional Experience

Alibaba Cloud Security WEBSHELL Text Detection
Sep 2022 – Dec 2022
Algorithm Analyst

  • Optimized the procedure of deploying BERT on Alibaba cloud platform for extensive Webshell detection based on team’s creative approaches..
  • Conducted detailed data analysis and summarization, performed ablation studies comparing local tests and platform scores to infer platform test data composition, strategically refined the deployed model.
  • Modified the algorithm through various methods, resulting in a final detection rate above 97%.

Community resource integration application during the COVID-19
Jan 2020 – Jun 2022
Institute of Software Hubei University, Wuhan, China
Systems architect

  • Aimed at difficulties in community management.
  • Provided a platform for community residents to assist each other.

Selected Publications

[1]Zeqing Qin, Zhiwei Ye*, et al. A Multimodal Biomedical Image Registration Method Based on an Improved Genetic Algorithm Inspired by Hybrid Breeding[C]//2021 International Conference on Systems, Man, and Cybernetics, IEEE SMC2021. (CCF: C, QUALIS: B2) paper

[2]Zesong Wang, Kui Xiao*, Zeqing Qin, et al. Extracting Prerequisites between Concepts from Online Course Introduction[C]//The 33rd International Conference on Software Engineering & Knowledge Engineering, SEKE2021. (acceptance rate: 23.5%, CCF: C, CORE: B, QUALIS: B1) paper

Competition and Awards

  • Representative of Undergraduate Graduates - Only one person among all graduates
  • Won the National Scholarship 2020
  • Won the Top 10% on Alibaba Tianchi Security algorithm Competition 2022
  • Won the First Prize in China Collegiate Computing Contest 2020
  • Won the Meritorious Winner in Mathematical Contest in Modeling 2020
  • Won the Second Prize in National Undergraduate Mathemetical Contest in Modeling 2020
  • Won the Third Prize on China Undergraduate Computer Design Contest 2019