The Unreasonable Difficulty of Time Series Forecasting

Preview

Preview I’ve been thinking recently about what makes time series forecasting problems so difficult compared to other sequence learning tasks or IID Machine Learning problems. First some motivating baselines illustrating the difficulty of the forecasting problem. I ran a series of models statistical (Naive, AutoARIMA, Theta, MSTL, Seasonal-Naive) linear/transformer neural (DLinear, NLinear, PatchTST) a gradient-boosted tree (LightGBM) zero-shot foundation models (Chronos, TimesFM, TTM) “AI” LLM prompted to continue numeric sequence (Claude Opus, Haiku) against the benchmark datasets m4_hourly,

saved by

• nicolas
• Akira Yoshiyama

related reading

• Against Time-Series Foundation Modelsshakoist.substack.com
• Time Series - From Analyzing the Past to Predicting the Future | Towards Data Sciencetowardsdatascience.com
• ML Blog - Everything you need to know about Time Series Forecastingmrtunguyen.github.io
• Zipfian grokking | Jasper Gilleyjagilley.github.io
• 2402.03885arxiv.org
• TimeGPT-1 | Hacker Newsnews.ycombinator.com
• An introduction to time series forecasting | InfoWorldinfoworld.com
• [2310.07820] Large Language Models Are Zero-Shot Time Series Forecastersarxiv.org
• The Illusion of Knowledgeoaktreecapital.com
• Khoa học dữ liệuphamdinhkhanh.github.io
• Introduction to Time Series Analysis — Time Series Analysis Handbookphdinds-aim.github.io
• GitHub - google-research/timesfm: TimesFM (Time Series Foundation Model) is a pretrained time-series foundation model developed by Google Research for time-series forecasting.github.com