%0 Conference Paper %F Oral %T Residual Integration Neural Network %+ Département Signal et Communications (IMT Atlantique - SC) %+ Institut Mediterrani d'Estudis Avancats (IMEDEA) %+ Lab-STICC_IMTA_CID_TOMS %A Ouala, Said %A Pascual, Ananda %A Fablet, Ronan %< avec comité de lecture %B ICASSP 2019 : IEEE International Conference on Acoustics, Speech and Signal Processing %C Brighton, United Kingdom %8 2019-05-12 %D 2019 %R 10.1109/ICASSP.2019.8683447 %K Dynamical systems %K Data-driven models %K Neural networks %K Forecasting %Z Statistics [stat]/Machine Learning [stat.ML]Conference papers %X In this work, we investigate residual neural network representations for the identification and forecasting of dynamical systems. We propose a novel architecture that jointly learns the dynamical model and the associated Runge-Kutta integration scheme. We demonstrate the relevance of the proposed architecture with respect to learning-based state-of-the-art approaches in the identification and forecasting of chaotic dynamics when provided with training data with low temporal sampling rates. %G English %2 https://imt-atlantique.hal.science/hal-02005399/document %2 https://imt-atlantique.hal.science/hal-02005399/file/Integration_Residual_Net%281%29.pdf %L hal-02005399 %U https://imt-atlantique.hal.science/hal-02005399 %~ UNIV-BREST %~ INSTITUT-TELECOM %~ CNRS %~ UNIV-UBS %~ ENIB %~ LAB-STICC %~ LAB-STICC_IMTA_CID_TOMS %~ IMTA_SC %~ LAB-STICC_IMTA %~ IMT-ATLANTIQUE %~ PRACOM %~ INSTITUTS-TELECOM %~ IMTA_MEE