SRDCNN: Strongly Regularized Deep Convolution Neural Network Architecture for Time-series Sensor Signal Classification Tasks

by   Arijit Ukil, et al.

Deep Neural Networks (DNN) have been successfully used to perform classification and regression tasks, particularly in computer vision based applications. Recently, owing to the widespread deployment of Internet of Things (IoT), we identify that the classification tasks for time series data, specifically from different sensors are of utmost importance. In this paper, we present SRDCNN: Strongly Regularized Deep Convolution Neural Network (DCNN) based deep architecture to perform time series classification tasks. The novelty of the proposed approach is that the network weights are regularized by both L1 and L2 norm penalties. Both of the regularization approaches jointly address the practical issues of smaller number of training instances, requirement of quicker training process, avoiding overfitting problem by incorporating sparsification of weight vectors as well as through controlling of weight values. We compare the proposed method (SRDCNN) with relevant state-of-the-art algorithms including different DNNs using publicly available time series classification benchmark (the UCR/UEA archive) time series datasets and demonstrate that the proposed method provides superior performance. We feel that SRDCNN warrants better generalization capability to the deep architecture by profoundly controlling the network parameters to combat the training instance insufficiency problem of real-life time series sensor signals.



There are no comments yet.


page 1

page 2

page 3

page 4


Handling Variable-Dimensional Time Series with Graph Neural Networks

Several applications of Internet of Things(IoT) technology involve captu...

Anomaly Detection And Classification In Time Series With Kervolutional Neural Networks

Recently, with the development of deep learning, end-to-end neural netwo...

Deep learning for time series classification: a review

Time Series Classification (TSC) is an important and challenging problem...

Learning Interpretable Shapelets for Time Series Classification through Adversarial Regularization

Times series classification can be successfully tackled by jointly learn...

Dynamic Time Warp Convolutional Networks

Where dealing with temporal sequences it is fair to assume that the same...

Detecting Gas Vapor Leaks Using Uncalibrated Sensors

Chemical and infra-red sensors generate distinct responses under similar...

Focusing on What is Relevant: Time-Series Learning and Understanding using Attention

This paper is a contribution towards interpretability of the deep learni...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.