A Random Matrix Approach to Echo-State Neural Networks

Romain Couillet, Gilles Wainrib, Hafiz Tiomoko Ali, Harry Sevi
Proceedings of The 33rd International Conference on Machine Learning, PMLR 48:517-525, 2016.

Abstract

Recurrent neural networks, especially in their linear version, have provided many qualitative insights on their performance under different configurations. This article provides, through a novel random matrix framework, the quantitative counterpart of these performance results, specifically in the case of echo-state networks. Beyond mere insights, our approach conveys a deeper understanding on the core mechanism under play for both training and testing.

Cite this Paper

BibTeX
@InProceedings{pmlr-v48-couillet16, title = {A Random Matrix Approach to Echo-State Neural Networks}, author = {Couillet, Romain and Wainrib, Gilles and Ali, Hafiz Tiomoko and Sevi, Harry}, booktitle = {Proceedings of The 33rd International Conference on Machine Learning}, pages = {517--525}, year = {2016}, editor = {Balcan, Maria Florina and Weinberger, Kilian Q.}, volume = {48}, series = {Proceedings of Machine Learning Research}, address = {New York, New York, USA}, month = {20--22 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v48/couillet16.pdf}, url = {https://proceedings.mlr.press/v48/couillet16.html}, abstract = {Recurrent neural networks, especially in their linear version, have provided many qualitative insights on their performance under different configurations. This article provides, through a novel random matrix framework, the quantitative counterpart of these performance results, specifically in the case of echo-state networks. Beyond mere insights, our approach conveys a deeper understanding on the core mechanism under play for both training and testing.} }
Endnote
%0 Conference Paper %T A Random Matrix Approach to Echo-State Neural Networks %A Romain Couillet %A Gilles Wainrib %A Hafiz Tiomoko Ali %A Harry Sevi %B Proceedings of The 33rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2016 %E Maria Florina Balcan %E Kilian Q. Weinberger %F pmlr-v48-couillet16 %I PMLR %P 517--525 %U https://proceedings.mlr.press/v48/couillet16.html %V 48 %X Recurrent neural networks, especially in their linear version, have provided many qualitative insights on their performance under different configurations. This article provides, through a novel random matrix framework, the quantitative counterpart of these performance results, specifically in the case of echo-state networks. Beyond mere insights, our approach conveys a deeper understanding on the core mechanism under play for both training and testing.
RIS
TY - CPAPER TI - A Random Matrix Approach to Echo-State Neural Networks AU - Romain Couillet AU - Gilles Wainrib AU - Hafiz Tiomoko Ali AU - Harry Sevi BT - Proceedings of The 33rd International Conference on Machine Learning DA - 2016/06/11 ED - Maria Florina Balcan ED - Kilian Q. Weinberger ID - pmlr-v48-couillet16 PB - PMLR DP - Proceedings of Machine Learning Research VL - 48 SP - 517 EP - 525 L1 - http://proceedings.mlr.press/v48/couillet16.pdf UR - https://proceedings.mlr.press/v48/couillet16.html AB - Recurrent neural networks, especially in their linear version, have provided many qualitative insights on their performance under different configurations. This article provides, through a novel random matrix framework, the quantitative counterpart of these performance results, specifically in the case of echo-state networks. Beyond mere insights, our approach conveys a deeper understanding on the core mechanism under play for both training and testing. ER -
APA
Couillet, R., Wainrib, G., Ali, H.T. & Sevi, H.. (2016). A Random Matrix Approach to Echo-State Neural Networks. Proceedings of The 33rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 48:517-525 Available from https://proceedings.mlr.press/v48/couillet16.html.

Related Material