Large Language Models for Time Series Forecasting

Abstract: In this talk, we will explore the application of large language models (LLMs) to time series forecasting.  We discuss how LLMs, traditionally used for natural language tasks, can be adapted to predict future values in sequential numerical data.  The presentation will cover techniques for encoding time series data into a format suitable for LLMs, fine-tuning strategies, and methods to interpret model outputs.  We will also compare the performance of LLM-based approaches with traditional statistical and machine learning forecasting methods across various domains and time scales.

Speaker’s Bio:  Kshitij Tayal's research interests lie in advancing machine learning and data science to tackle real-world problems with significant societal and scientific impacts.  He has published his research at major computer science conferences, including the International Conference on Data Mining, the SIAM Conference on Data Mining, the International Conference on Knowledge Discovery and Data Mining, the Conference on Computational Linguistics, Language Resources and Evaluation, and the ACM Conference on Intelligent User Interfaces.  He is working to build a novel reprogramming framework that can repurpose LLMs for scientific use cases by aligning natural language modalities to scientific data.