Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Dirichlet-Guided Group Forecasting for Alleviating Over-smoothing in Time Series Forecasting

About

Time series forecasting often suffers from over-smoothing, especially when future dynamics are multi-modal. Forecasts may follow the coarse trend of the observed future, but fail to preserve sharp changes, oscillations, turning points, and regime transitions that define plausible dynamic evolution. In this work, we revisit over-smoothing from the perspective of latent dynamical mode compression: under partial observation and single-realization supervision, multiple plausible future modes can be weakened, merged, or averaged during forecasting. Based on this view, we propose Dirichlet-Guided Group Forecasting (DGF), a mode-preserving forecasting framework that explicitly models multiple mode-conditioned predictive distributions and uncertainty over their selection probabilities. DGF uses a Dirichlet-guided hierarchical sampling mechanism and reward-based optimization to encourage forecasts that are accurate, dynamically consistent, and mode-distinct. Extensive experiments on real-world forecasting benchmarks show that DGF reduces over-smoothing while improving forecasting accuracy, diversity, and dynamical consistency.

Xingyu Zhang, Jingyao Wang, Xin Yu, Zeen Song, Jianqi Zhang, Changwen Zheng, Wenwen Qiang• 2026

Related benchmarks

TaskDatasetResultRank
Long-term forecastingETTh1
MSE0.395
500
Long-term time-series forecastingETTm2
MSE0.234
479
Long-term forecastingETTm1
MSE0.317
452
Long-term forecastingETTh2
MSE0.337
376
Long-term time-series forecastingWeather (test)
MSE0.28
240
Long-term forecastingETTm1
MAE0.322
86
Long-term forecastingWeather
MAE0.167
58
Long Sequence ForecastingElectricity
MAE0.219
30
Long-term forecastingETTm1 (test)
MSE0.265
29
Time Series ForecastingTraffic
Best MSE0.247
11
Showing 10 of 11 rows

Other info

Follow for update