TY - GEN
T1 - Detecting hazardous events from sequential data with multilayer architectures
AU - Kurach, Karol
AU - Pawlowski, Krzysztof
N1 - Publisher Copyright:
Copyright © 2015 for the individual papers by the papers' authors.
Copyright:
Copyright 2016 Elsevier B.V., All rights reserved.
PY - 2015
Y1 - 2015
N2 - Multivariate time series data play an important role in many domains, including real-time monitoring systems. In this paper, we focus on multilayer neural architectures that are capable of learning high level representations from raw data. This includes our previous solution based on Recurrent Neural Networks with Long Short-Term Memory (LSTM) cells. We build upon this work and present improved methods that aim to achieve higher prediction quality and better generalization to other similar tasks. We apply new deep neural architectures, minimize feature engineering and explore different ways of model selection. In particular, our focus on architectures includes networks with attention mechanism and convolutional networks. We tackle overfitting challenges in a presence of concept drift.
AB - Multivariate time series data play an important role in many domains, including real-time monitoring systems. In this paper, we focus on multilayer neural architectures that are capable of learning high level representations from raw data. This includes our previous solution based on Recurrent Neural Networks with Long Short-Term Memory (LSTM) cells. We build upon this work and present improved methods that aim to achieve higher prediction quality and better generalization to other similar tasks. We apply new deep neural architectures, minimize feature engineering and explore different ways of model selection. In particular, our focus on architectures includes networks with attention mechanism and convolutional networks. We tackle overfitting challenges in a presence of concept drift.
KW - Deep learning
KW - Ensemble methods
KW - Recurrent neural networks
KW - Time series forecasting
UR - http://www.scopus.com/inward/record.url?scp=84955451124&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84955451124&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:84955451124
T3 - CEUR Workshop Proceedings
SP - 1
EP - 10
BT - Concurrency, Specification and Programming - 24th InternationalWorkshop, CS and P 2015, Proceedings
A2 - Suraj, Zbigniew
A2 - Czaja, Ludwik
A2 - Czaja, Ludwik
PB - CEUR-WS
T2 - 24th International Workshop on Concurrency, Specification and Programming, CS and P 2015
Y2 - 28 September 2015 through 30 September 2015
ER -