𝙃𝙚𝙡𝙡𝙤, 𝙄'𝙢 Dai Ma
A Computer Science undergrad at Sun Yat-sen University, passionate about Text Analytics, Data Mining, and Recommendation Systems.
> About Me
I am a dedicated and self-driven student with a solid grounding in algorithms, data structures, and software engineering. My academic and project experiences have equipped me with strong problem-solving skills, enabling me to lead multiple deep learning projects from conception to deployment. I am deeply passionate about academic research and aspire to pursue a Ph.D. to contribute to the advancement of text analysis and data mining.
- 📧 Email: madai@mail2.sysu.edu.cn
- 🐱 GitHub: KDAIer
- 📞 Phone: +86 198 6489 5373
> Research Experience
Research Assistant, HKUST
Supervisor: Prof. Zixuan Yuan | Apr. 2025 – Present
Focused on Multimodal Large Language Models (LLMs). My work involves developing and evaluating Retrieval-Augmented Generation (RAG) systems to enhance model performance by integrating external knowledge, aiming to improve reasoning and reduce hallucinations in complex question-answering tasks.
> Key Projects
Retrieval-Augmented Generation (RAG) System
Project Lead & Developer
Developed a comprehensive RAG system designed to enhance large language models by retrieving relevant, up-to-date information from external knowledge bases before generating responses. This project explores advanced indexing and retrieval strategies to improve answer accuracy.
Risk Control Driven Adversarial Spam Text Detection
Project Leader
Engineered an end-to-end spam detection pipeline. A key innovation was a character similarity network designed to identify and neutralize adversarial text variants in Chinese, which significantly boosted recall and F1-score in risk control scenarios.
> Other Projects & Coursework
CS-Review: A Study Guide for SYSU CS Undergrads
Creator & Maintainer
Initiated and maintained a knowledge-sharing repository for computer science students at Sun Yat-sen University. This project consolidates study notes, exam materials, and learning resources to foster a collaborative academic environment.
Maximizing National Performance in the Olympic Games
Project Leader, COMAP MCM
Built a medal prediction system combining XGBoost, BP neural network, and multivariate regression, achieving 90% F1 for medal types and forecasting future results.
Image Classification: Optimization and Model Comparison
Project Leader
Enhanced CIFAR-10 classification by comparing Softmax, MLP, and CNN in PyTorch. Tuned various optimizers, where Adam achieved 30% faster convergence and a CNN accuracy of 70%.
> Awards & Honors
- Meritorious Winner, Mathematical Contest in Modeling (MCM/ICM) - 2025
- Bronze Prize, Sun Yat-sen University Programming Contest - 2024
- Third Prize, University-level Scholarship, Sun Yat-sen University - 2024
- Third Prize, Guangdong Province, National Undergraduate Mathematics Competition - 2023
> My GitHub Stats 📈
> My Tech Stack
