A fast noise filtering algorithm for time series prediction using recurrent neural networks

Recent research demonstrate that prediction of time series by recurrent neural networks (RNNs) based on the noisy input generates a smooth anticipated trajectory. We examine the internal dynamics of RNNs and establish a set of conditions required for such behavior. Based on this analysis we propose a new approximate algorithm and show that it significantly speeds up the predictive process without loss of accuracy.

cs.LG math.DS stat.ML Neural Networks and Applications Time Series Analysis and Forecasting Model Reduction and Neural Networks Computer Science Mathematics

View graph Source

Paper

References (13)

01Language Models are Few-Shot LearnersTom B. Brown, Benjamin Mann, Nick Ryder et al.2020 · Neural Information Processing Systems · 61k citations In Library

02Short note on the behavior of recurrent neural network for noisy dynamical systemK. Yeo2019 · arXiv.org · 15 citations In Library

03Moving Objects Analytics: Survey on Future Location & Trajectory Prediction MethodsHarris V. Georgiou, Sophia Karagiorgou, Y. Kontoulis et al.2018 · arXiv.org · 38 citations In Library

04Data-driven forecasting of high-dimensional chaotic systems with long short-term memory networksPantelis R. Vlachas, Wonmin Byeon, Z. Y. Wan et al.2018 · Proceedings of the Royal Society A · 563 citations In Library

05Attention is All you NeedAshish Vaswani, Noam Shazeer, Niki Parmar et al.2017 · Neural Information Processing Systems · 187k citations In Library

06A New Approach for Chaotic Time Series Prediction Using Recurrent Neural NetworkQinghai Li, Rui-Chang Lin2016 · 23 citations

07Long Short-Term MemorySepp Hochreiter, J. Schmidhuber1997 · Neural Computation · 108k citations

08Learning Chaotic Dynamics using Tensor Recurrent Neural NetworksRose Yu, Stephan Zheng, Yan Liu2017 · 22 citations

09Language models are few-shot learnenrs, 2020, arXiv:2005.14165v4 [cs.CL2005

10Byeon et al, Data-driven forecasting of high-dimensional chaotic systems with long short-term memory networks2017 · Proc. R. Soc. A

11Attention is all you need, 2017, arXiv:1706.03762v5 [cs.CL2017

12Long-shotr term memory, Neural1997 · Comput., 1997,

Scroll for more · 1 remaining

Similar papers

It's a super deal -- train recurrent network on noisy data and get smooth prediction freeBoris Rubinstein2022 · arXiv (Cornell University) In Library

Probabilistic, Recurrent, Fuzzy Neural Network for Processing Noisy Time-Series DataYong Li, Richard Gault, T.M. McGinnity2021 · IEEE Transactions on Neural Networks and Learning Systems · 37 citations In Library

Noisy Recurrent Neural NetworksS. H. Lim, N. Benjamin Erichson, Liam Hodgkinson et al.2021 · Neural Information Processing Systems · 74 citations In Library

Efficient Time-Series Approximation with Linear Recurrent Neural Networks: Architecture Learning and Predictive PowerFrieder Stolzenburg, Sandra Litz, O. Michael et al.2018 · Neural computing & applications (Print) · 8 citations In Library

Analysis of Time Series Prediction using Recurrent Neural NetworksGaurav Yadav, Richa Vasuja2019 · International Journal of Computer Applications · 4 citations In Library

A fast memoryless predictive algorithm in a chain of recurrent neural networksBoris Rubinstein2020 · arXiv (Cornell University) · 2 citations In Library

Efficient time-series approximation with linear recurrent neural networks: architecture learning and predictive powerFrieder Stolzenburg, Sandra Litz, Olivia Michael et al.2025 · Neural Computing and Applications In Library

Approximate nonlinear filtering with a recurrent neural networkA. Kutschireiter, S. C. Surace, Henning Sprekeler et al.2015 · BMC Neuroscience · 1 citation In Library

Robust Learning of Recurrent Neural Networks in Presence of Exogenous NoiseArash Amini, Guangyi Liu, Nader Motee2021 · arXiv (Cornell University) · 5 citations In Library

Recurrent Neural Networks for Time Series Forecasting: Current Status and Future DirectionsHansika Hewamalage, C. Bergmeir, Kasun Bandara2019 · International Journal of Forecasting · 1.3k citations In Library

Source of record

arXiv

44B

Welcome.

This panel sticks with you. Pick anything below and it opens right beside it, so you can dig through 60,000-plus records without ever losing your spot here.

Travel 44B

LibraryPapers, policy, standards, statute LabsEvery organization building AI ModelsIntelligence, price, and speed BenchmarksThe evaluation catalog IncidentsAI failures and harms SearchOne field across everything DashboardThe state of AI accountability in NY 44B RegistryThe Art. 44-B compliance portal