aiqbench analysis
This arXiv paper proposes a neural method that propagates whole predictive distributions through recursive forecasting, using kernel mean embeddings for inputs and mixture density networks for outputs, instead of approximating uncertainty with ensembles. The evidence is a proof-of-concept on the Lorenz63 chaotic system, so it shows feasibility on a low-dimensional toy problem rather than performance on high-dimensional real-world forecasting. Readers interested in uncertainty quantification for dynamical systems should watch whether the approach scales beyond this setting.
A Distribution-to-Distribution Neural Probabilistic Forecasting Framework for Dynamical Systems
arXiv:2603.25370v2 Announce Type: replace-cross Abstract: Probabilistic forecasts provide a principled representation of forecast uncertainty through predictive distributions. Directly propagating predictive distributions with conventional physical models requires solving distributional evolution equations, a task well-known to be computationally prohibitive for high-dimensional nonlinear systems. Consequently, multi-step predictive distributions are commonly approximated using ensembles or sampled trajectories. A novel distribution-to-distribution (D2D) neural probabilistic forecasting framework is proposed with compatible distributional encoding and decoding interfaces around a replaceable neural forecasting module. Kernel mean embeddings represent input distributions, while mixture density networks parameterise outputs, enabling recursive multi-step forecasting in distributional form. The framework is demonstrated on the Lorenz63 system. Results show that the D2D model captures nontrivial distributional evolution, produces skillful probabilistic forecasts without explicit ensemble simulation, and remains competitive with a naively initialised perfect-model benchmark. This provides proof-of-concept evidence that predictive distributions can be recursively propagated through a neural forecasting model and evaluated in terms of probabilistic forecast performance.
Abstract:Probabilistic forecasts provide a principled representation of forecast uncertainty through predictive distributions. Directly propagating predictive distributions with conventional physical models requires solving distributional evolution equations, a task well-known to be computationally prohibitive for high-dimensional nonlinear systems. Consequently, multi-step predictive distributions are commonly approximated using ensembles or sampled trajectories. A novel distribution-to-distribution (D2D) neural probabilistic forecasting framework is proposed with compatible distributional encoding and decoding interfaces around a replaceable neural forecasting module. Kernel mean embeddings represent input distributions, while mixture density networks parameterise outputs, enabling recursive multi-step forecasting in distributional form. The framework is demonstrated on the Lorenz63 system. Results show that the D2D model captures nontrivial distributional evolution, produces skillful probabilistic forecasts without explicit ensemble simulation, and remains competitive with a naively initialised perfect-model benchmark. This provides proof-of-concept evidence that predictive distributions can be recursively propagated through a neural forecasting model and evaluated in terms of probabilistic forecast performance.
Machine Learning (stat.ML); Machine Learning (cs.LG)
I.2.6; I.6.5; G.3
arXiv:2603.25370 [stat.ML]
(or arXiv:2603.25370v2 [stat.ML] for this version)
https://doi.org/10.48550/arXiv.2603.25370 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 247 words
Related coverage
News
Quantum Computing Report · Jul 25, 2026
Variational Augmented Invertible Koopman Autoencoder for probabilistic time series forecasting
arXiv · Sep 30, 2026
DynG-Diff: A State-Aware Dynamic Guidance Diffusion Framework for Probabilistic Time Series Forecasting
arXiv · Sep 3, 2026
The Entangled Dimensions of Decision Theory
LessWrong · Jul 30, 2026