- 🔭 Current Focus: Pursuing a Master's in CS at TU Dresden. I am interested in Deep learning, AI applications, and LLM.
- 🧠 Research Interests:
- Large Language Models: Specialized in PEFT (LoRA/QLoRA) and domain-specific fine-tuning.
- RAG Optimization: Enhancing retrieval accuracy and context injection for data-driven systems.
- Deep Learning: Architecting neural networks for predictive modeling and computer vision.
- 📍 Based in: Dresden, Germany 🇩🇪
I am passionate about solving real-world problems through predictive modeling.
- [ACM Digital Library] Using the Temporal-Trajectory-based KNN to Predict Human Mobility Patterns
- Ranked Top 10 / 100 globally in the Human Mobility Prediction Challenge 2024.
- IEEE-ICASI 2025 Personalized Travel Recommendation Systems
- Proposed an N-gram-based system to solve the cold-start problem in tourism.
- IEEE-ICASI 2025 Predictive Modeling of Traffic Accidents in Taiwan
- Designed a GPT-based model to address the zero-inflation problem in traffic datasets.
| Category | Tools & Languages |
|---|---|
| Languages | Python, C, Java, SQL |
| AI / ML | PyTorch, TensorFlow, Keras, Scikit-learn, Pandas, NumPy |
| LLM Tech | Llama Fine-tuning (LoRA), RAG (Retrieval-Augmented Generation) |
| Developer Tools | Git, GitHub, Linux (HPC environments) |
- [LLM_cultral_qa] - Fine-tuning Llama 8B using LoRA and implementing RAG for specialized cultural databases.
- [Intersection Recognition] - CNN-based pipeline for identifying high-risk intersections using satellite imagery.
- [Human Mobility KNN] - High-efficiency trajectory prediction algorithm.
- 📧 Email: qazxcvb1425@gmail.com
- 📱 Phone: +49 15141349697
- 🌍 Languages: Mandarin (Native), English (IELTS C1), German (A1 Learning)