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

EnTransformer: A Deep Generative Transformer for Multivariate Probabilistic Forecasting

About

Reliable uncertainty quantification is critical in multivariate time series forecasting problems arising in domains such as energy systems and transportation networks, among many others. Although Transformer-based architectures have recently achieved strong performance for sequence modeling, most probabilistic forecasting approaches rely on restrictive parametric likelihoods or quantile-based objectives. They can struggle to capture complex joint predictive distributions across multiple correlated time series. This work proposes EnTransformer, a deep generative forecasting framework that integrates engression, a stochastic learning paradigm for modeling conditional distributions, with the expressive sequence modeling capabilities of Transformers. The proposed approach injects stochastic noise into the model representation and optimizes an energy-based scoring objective to directly learn the conditional predictive distribution without imposing parametric assumptions. This design enables EnTransformer to generate coherent multivariate forecast trajectories while preserving Transformers' capacity to effectively model long-range temporal dependencies and cross-series interactions. We evaluate our proposed EnTransformer on several widely used benchmarks for multivariate probabilistic forecasting, including Electricity, Traffic, Solar, Taxi, KDD-cup, and Wikipedia datasets. Experimental results demonstrate that EnTransformer produces well-calibrated probabilistic forecasts and consistently outperforms the benchmark models.

Rajdeep Pathak, Rahul Goswami, Madhurima Panja, Palash Ghosh, Tanujit Chakraborty• 2026

Related benchmarks

TaskDatasetResultRank
ForecastingKDDCUP
CRPS0.2468
22
Multivariate probabilistic forecastingsolar
CRPS-sum0.2421
12
Time Series Forecastingsolar
NRMSE Sum0.3129
10
Time Series ForecastingElectricity
NRMSE Sum2.78
10
Time Series Forecastingtaxi
NRMSE Sum0.1516
10
Time Series ForecastingTraffic
NRMSE Sum0.0789
10
Time Series ForecastingKDD-cup
NRMSE (sum)0.3094
10
Multivariate probabilistic forecastingElectricity
CRPS Sum0.0216
9
Probabilistic Forecastingtaxi
CRPS0.119
9
Time Series ForecastingWikipedia
NRMSE Sum0.083
9
Showing 10 of 12 rows

Other info

Follow for update