Projects
A collection of my key projects in AI, machine learning, and financial market analysis.
Featured Projects
PACO - Pink Ant Colony Optimization
A novel implementation of Ant Colony Optimization (ACO) algorithm for solving shortest path and Traveling Salesman Problems (TSP). PACO incorporates adaptive gradient descent, elitist strategy, MaxMin boundaries, and parallel processing to enhance solution quality while reducing computational cost.
Key Contributions:
- Developed an optimized algorithm combining gradient descent and pheromone updates
- Implemented parallelization techniques for significant performance improvement
- Applied advanced techniques like Floyd Warshall preprocessing and 2-opt local search
Mitigation of Large Language Model (LLM) Hallucination
Addressing hallucination in LLMs using Fine-Tuning (LoRA & QLoRA) and Self-Refine techniques, applied to LLaMA 3 8B Instruct in the medical domain. Our model outperforms the base model in MELT-HALT and HaluBench benchmarks.
Key Contributions:
- Led research and project management
- Developed and fine-tuned models locally and on high-performance computing infrastructure
- Worked on LoRA, QLoRA, prompt engineering, and self-refinement techniques
Melbourne House Market Analysis using ML
A machine learning project analyzing Melbourne’s house market for investors. Deployed models via a full-stack web application using React (Front-End) and FastAPI (Back-End).
Key Contributions:
- Collected and processed real estate data
- Implemented Random Forest Regression/Classification and Clustering techniques
- Developed and deployed ML models through a RESTful API
Other Projects
Vietnamese Blog Titles Classification using Transformers
A deep learning project to classify Vietnamese blog titles using Transformer models.
Key Contributions:
- Implemented PyTorch-based Transformer models
- Explored advanced optimization techniques: ROBE, Mixture of Experts (MoE), synthetic data generation
Market Intelligence Analysis for Commodity Futures Investment
A real-world industry project supporting commodity futures trading (gold, oil) using ML/DL. Analyzed geopolitical, economic, and financial factors.
Key Contributions:
- Developed predictive models with Transformers, LSTMs, GRUs, and CNNs
- Conducted feature engineering for complex market analysis
- Applied interdisciplinary approach combining finance and AI
Get in Touch
Interested in collaborating on a project? Have questions about my work? Feel free to contact me for more information.