👋Yile Li(李怡乐)

Developed a deep and comprehensive understanding of foundational deep learning techniques through theoretical study and extensive model reproduction.

Conducted a comprehensive literature review on state-of-the-art recommender systems, focusing on bundle generation methods. Selected and replicated baseline models to ensure a solid foundation for further research.

Developed a bundle generation pipeline that integrates GNN's structural-awareness with LLM-enhanced data embeddings to incorporate open-world knowledge, using noise-based training to improve robustness and reduce the impact of low-quality data. Currently refining the model for enhanced performance.

Led front-end and back-end development using Vue.js and Flask.

Fine-tuned a large model with LoRA methods for its application on brand reputation analysis.

Conducted preliminary research to develop a brand reputation index using AHP and Entropy Method.

Collaborated with the National Institute of Advertising, providing technical support for the 2025 China Advertising Great Wall Award.

Contributed to the development of the digital museum platform using Vue.js.

Implemented the HLS protocol encrypt streaming video, generating M3U8 index files for secure content transmission.