paper-with-me

홈 › Papers

Probabilistic Recurrent State-Space Models

2018-01-31 · ICML 2018 7 · Andreas Doerr, Christian Daniel, Martin Schiegg, Duy Nguyen-Tuong, Stefan Schaal, Marc Toussaint, Sebastian Trimpe

State-space models (SSMs) are a highly expressive model class for learning patterns in time series data and for system identification. Deterministic versions of SSMs (e.g. LSTMs) proved extremely successful in modeling complex time series data. Fully probabilistic SSMs, however, are often found hard to train, even for smaller problems. To overcome this limitation, we propose a novel model formulation and a scalable training algorithm based on doubly stochastic variational inference and Gaussian processes. In contrast to existing work, the proposed variational approximation allows one to fully capture the latent state temporal correlations. These correlations are the key to robust training. The effectiveness of the proposed PR-SSM is evaluated on a set of real-world benchmark datasets in comparison to state-of-the-art probabilistic model learning methods. Scalability and robustness are demonstrated on a high dimensional problem.

📄 PDF Abstract BibTeX arXiv:1801.10395

Code (4)

andreasdoerr/PR-SSM 공식 구현 tf
boschresearch/PR-SSM tf
zhidilin/gpssmproj pytorch
zhidilin/odgpssm pytorch

Tasks

Gaussian ProcessesState Space ModelsTime SeriesTime Series AnalysisVariational Inference

Similar Papers 제목 키워드 기반

Uncertainty Representations in State-Space Layers for Deep Reinforcement Learning under Partial Observability

2024-09-25 · Carlos E. Luis, Alessandro G. Bottero, Julia Vinogradska, Felix Berkenkamp 외

Optimal decision-making under partial observability requires reasoning about the uncertainty of the environment's hidden state. However, most reinforcement learning architectures handle partial observability with sequenc…

Decision MakingDeep Reinforcement Learningreinforcement-learningReinforcement Learning+1

Probabilistic Verification of Recurrent Neural Networks for Single and Multi-Agent Reinforcement Learning

2026-05-14 · Luca Marzari, Enrico Marchesini arxiv

History-dependent policies induced by recurrent neural networks (RNNs) rely on latent hidden state dynamics, making verification in partially observable reinforcement learning (RL) challenging. Existing RNN verification …

Multi-agent Reinforcement Learning

Anomaly Detection for Non-stationary Time Series using Recurrent Wavelet Probabilistic Neural Network

2025-05-16 · Pu Yang, J. A. Barria

In this paper, an unsupervised Recurrent Wavelet Probabilistic Neural Network (RWPNN) is proposed, which aims at detecting anomalies in non-stationary environments by modelling the temporal features using a nonparametric…

Anomaly DetectionDecoderDensity EstimationTime Series+1

Deep Probabilistic Time Series Forecasting using Augmented Recurrent Input for Dynamic Systems

2021-06-03 · Haitao Liu, Changjun Liu, Xiaomo Jiang, Xudong Chen 외

The demand of probabilistic time series forecasting has been recently raised in various dynamic system scenarios, for example, system identification and prognostic and health management of machines. To this end, we combi…

DecoderManagementProbabilistic Time Series ForecastingTime Series+2

Learning Interpretable Deep State Space Model for Probabilistic Time Series Forecasting

2021-01-31 · Longyuan Li, Junchi Yan, Xiaokang Yang, Yaohui Jin

Probabilistic time series forecasting involves estimating the distribution of future based on its history, which is essential for risk management in downstream decision-making. We propose a deep state space model for pro…

Decision MakingManagementProbabilistic Time Series ForecastingTime Series+2