M-4 Competition

The M4 Competition is a benchmark dataset that contains a diverse set of time series data used to evaluate the forecasting accuracy of various time series models.

Papers using M-4 Competition (11)

Rethinking Nonstationarity in Time Series: A Deterministic Trend Perspective2025Recurrent Neural Networks for Time Series Forecasting: Current Status and Future Directions2019 · 1,196 citesTwo-Step Meta-Learning for Time-Series Forecasting Ensemble2020 · 16 citesThe uncertainty estimation of feature-based forecast combinations2019 · 3 citesFor2For: Learning to forecast from forecasts2020 · 2 citesHERMES: Hybrid Error-corrector Model with inclusion of External Signals for nonstationary fashion time series2022 · 2 citesDesigning Time-Series Models With Hypernetworks & Adversarial Portfolios2024 · 2 citesTime Series Analysis by State Space Learning2024 · 1 citesTo aggregate or not to aggregate: Forecasting of finite autocorrelated demand2021Optimizing accuracy and diversity: a multi-task approach to forecast combinations2023Infinite forecast combinations based on Dirichlet process2023