Echo State Hoeffding Tree Learning

Diego Marron, Jesse Read, Albert Bifet, Talel Abdessalem, Eduard Ayguade, José Herrero
; Proceedings of The 8th Asian Conference on Machine Learning, PMLR 63:382-397, 2016.

Abstract

Nowadays, real-time classification of Big Data streams is becoming essential in a variety of application domains. While decision trees are powerful and easy-to-deploy approaches for accurate and fast learning from data streams, they are unable to capture the strong temporal dependences typically present in the input data. Recurrent Neural Networks are an alternative solution that include an internal memory to capture these temporal dependences; however their training is computationally very expensive and with slow convergence, requiring a large number of hyper-parameters to tune. Reservoir Computing was proposed to reduce the computation requirements of the training phase but still include a feed-forward layer which requires a large number of parameters to tune. In this work we propose a novel architecture for real-time classification based on the combination of a Reservoir and a decision tree. This combination reduces the number of hyper-parameters while still maintaining the good temporal properties of recurrent neural networks. The capabilities of the proposed architecture to learn some typical string-based functions with strong temporal dependences are evaluated in the paper. We show how the new architecture is able to incrementally learn these functions in real-time with fast adaptation to unknown sequences. And we study the influence of the reduced number of hyper-parameters in the behaviour of the proposed solution.

Cite this Paper


BibTeX
@InProceedings{pmlr-v63-Marron87, title = {Echo State Hoeffding Tree Learning}, author = {Diego Marron and Jesse Read and Albert Bifet and Talel Abdessalem and Eduard Ayguade and José Herrero}, booktitle = {Proceedings of The 8th Asian Conference on Machine Learning}, pages = {382--397}, year = {2016}, editor = {Robert J. Durrant and Kee-Eung Kim}, volume = {63}, series = {Proceedings of Machine Learning Research}, address = {The University of Waikato, Hamilton, New Zealand}, month = {16--18 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v63/Marron87.pdf}, url = {http://proceedings.mlr.press/v63/Marron87.html}, abstract = {Nowadays, real-time classification of Big Data streams is becoming essential in a variety of application domains. While decision trees are powerful and easy-to-deploy approaches for accurate and fast learning from data streams, they are unable to capture the strong temporal dependences typically present in the input data. Recurrent Neural Networks are an alternative solution that include an internal memory to capture these temporal dependences; however their training is computationally very expensive and with slow convergence, requiring a large number of hyper-parameters to tune. Reservoir Computing was proposed to reduce the computation requirements of the training phase but still include a feed-forward layer which requires a large number of parameters to tune. In this work we propose a novel architecture for real-time classification based on the combination of a Reservoir and a decision tree. This combination reduces the number of hyper-parameters while still maintaining the good temporal properties of recurrent neural networks. The capabilities of the proposed architecture to learn some typical string-based functions with strong temporal dependences are evaluated in the paper. We show how the new architecture is able to incrementally learn these functions in real-time with fast adaptation to unknown sequences. And we study the influence of the reduced number of hyper-parameters in the behaviour of the proposed solution.} }
Endnote
%0 Conference Paper %T Echo State Hoeffding Tree Learning %A Diego Marron %A Jesse Read %A Albert Bifet %A Talel Abdessalem %A Eduard Ayguade %A José Herrero %B Proceedings of The 8th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2016 %E Robert J. Durrant %E Kee-Eung Kim %F pmlr-v63-Marron87 %I PMLR %J Proceedings of Machine Learning Research %P 382--397 %U http://proceedings.mlr.press %V 63 %W PMLR %X Nowadays, real-time classification of Big Data streams is becoming essential in a variety of application domains. While decision trees are powerful and easy-to-deploy approaches for accurate and fast learning from data streams, they are unable to capture the strong temporal dependences typically present in the input data. Recurrent Neural Networks are an alternative solution that include an internal memory to capture these temporal dependences; however their training is computationally very expensive and with slow convergence, requiring a large number of hyper-parameters to tune. Reservoir Computing was proposed to reduce the computation requirements of the training phase but still include a feed-forward layer which requires a large number of parameters to tune. In this work we propose a novel architecture for real-time classification based on the combination of a Reservoir and a decision tree. This combination reduces the number of hyper-parameters while still maintaining the good temporal properties of recurrent neural networks. The capabilities of the proposed architecture to learn some typical string-based functions with strong temporal dependences are evaluated in the paper. We show how the new architecture is able to incrementally learn these functions in real-time with fast adaptation to unknown sequences. And we study the influence of the reduced number of hyper-parameters in the behaviour of the proposed solution.
RIS
TY - CPAPER TI - Echo State Hoeffding Tree Learning AU - Diego Marron AU - Jesse Read AU - Albert Bifet AU - Talel Abdessalem AU - Eduard Ayguade AU - José Herrero BT - Proceedings of The 8th Asian Conference on Machine Learning PY - 2016/11/20 DA - 2016/11/20 ED - Robert J. Durrant ED - Kee-Eung Kim ID - pmlr-v63-Marron87 PB - PMLR SP - 382 DP - PMLR EP - 397 L1 - http://proceedings.mlr.press/v63/Marron87.pdf UR - http://proceedings.mlr.press/v63/Marron87.html AB - Nowadays, real-time classification of Big Data streams is becoming essential in a variety of application domains. While decision trees are powerful and easy-to-deploy approaches for accurate and fast learning from data streams, they are unable to capture the strong temporal dependences typically present in the input data. Recurrent Neural Networks are an alternative solution that include an internal memory to capture these temporal dependences; however their training is computationally very expensive and with slow convergence, requiring a large number of hyper-parameters to tune. Reservoir Computing was proposed to reduce the computation requirements of the training phase but still include a feed-forward layer which requires a large number of parameters to tune. In this work we propose a novel architecture for real-time classification based on the combination of a Reservoir and a decision tree. This combination reduces the number of hyper-parameters while still maintaining the good temporal properties of recurrent neural networks. The capabilities of the proposed architecture to learn some typical string-based functions with strong temporal dependences are evaluated in the paper. We show how the new architecture is able to incrementally learn these functions in real-time with fast adaptation to unknown sequences. And we study the influence of the reduced number of hyper-parameters in the behaviour of the proposed solution. ER -
APA
Marron, D., Read, J., Bifet, A., Abdessalem, T., Ayguade, E. & Herrero, J.. (2016). Echo State Hoeffding Tree Learning. Proceedings of The 8th Asian Conference on Machine Learning, in PMLR 63:382-397

Related Material