종합설계프로젝트 사전 논문리뷰 코드
- Dataset 수집
- 6가지 논문에서 소개하는 코드 구현 (baseline models -embedding- 구현 / Paper List 참조)
- 6가지 baseline model (embedding) 에 대한 6가지 downstream task 실행 (Downstream Task List 참조)
- 결과 정리 및 시각화 (Table & Figure)
- aLLM4TS: Multi-Path Prediction - Adapting LLMs for Time Series Representation Learning
- Chronos (Will be added!)
- Merit: Multi-Agent Collaboration for Unsuperevised Time Series Representation Learning
- TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment
- TNC (Will be added!)
- TS2Vec: Towards Universal Representation of Time Series
※ 참고논문: 시계열 데이터 분석에 LLM을 이용한 모델
- TimekD: Efficient Multivariate Time Series Forecasting via Calibrated Language Models with Privileged Knowledge Distillation
- Anomaly detection
- Clustering
- Classification
- Forecasting
- Representation learning
- Transfer learning
- Intel Core i7-9700K @3.6GHz (8 Physical core / L1 Cache 64KB, L2 Cache 256KB per Core)
- DDR4 64GB
- NVIDIA GeForce RTX3090 24GB
- NVIDIA GeForce RTX3080Ti 12GB