Cost-sensitive convolutional neural networks for imbalanced time series classification

Geng, Yingsan; Luo, Xinyu

doi:10.3233/ida-183831

Cited by 36 publications

(23 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In Wang et al (2017b); Geng and Luo (2018), an MLP was designed to learn from scratch a discriminative time series classifier. The problem with an MLP approach is that temporal information is lost and the features learned are no longer time-invariant.…”

Section: Discriminative Modelsmentioning

confidence: 99%

Deep learning for time series classification: a review

Fawaz

Forestier

Weber

et al. 2019

Data Min Knowl Disc

2,062

712

View full text Add to dashboard Cite

Time Series Classification (TSC) is an important and challenging problem in data mining. With the increase of time series data availability, hundreds of TSC algorithms have been proposed. Among these methods, only a few have considered Deep Neural Networks (DNNs) to perform this task. This is surprising as deep learning has seen very successful applications in the last years. DNNs have indeed revolutionized the field of computer vision especially with the advent of novel deeper architectures such as Residual and Convolutional Neural Networks. Apart from images, sequential data such as text and audio can also be processed with DNNs to reach state-of-the-art performance for document classification and speech recognition. In this article, we study the current state-ofthe-art performance of deep learning algorithms for TSC by presenting an empirical study of the most recent DNN architectures for TSC. We give an overview of the most successful deep learning applications in various time series domains under a unified taxonomy of DNNs for TSC. We also provide an open source deep learning framework to the TSC community where we implemented each of the compared approaches and evaluated them on a univariate TSC benchmark (the UCR/UEA archive) and 12 multivariate time series datasets. By training 8,730 deep learning models on 97 time series datasets, we propose the most exhaustive study of DNNs for TSC to date.

show abstract

Section: Discriminative Modelsmentioning

confidence: 99%

Deep learning for time series classification: a review

Fawaz

Forestier

Weber

et al. 2019

Data Min Knowl Disc

2,062

712

View full text Add to dashboard Cite

show abstract

“…Among them, cost-sensitive learning is a common method, which avoids adding virtual samples which may destroy the prior distributions of variables. It has been successfully integrated in may many methods, such as support vector machines [17], neural networks [41], Bayesian methods [42], etc.…”

Section: Cost Sensitive Learning For Blsmentioning

confidence: 99%

An Efficient Anomaly Detection for High-Speed Train Braking System Using Broad Learning System

Wang

Liu

2021

IEEE Access

View full text Add to dashboard Cite

For a high-speed train, the braking system plays an essential role in safe transportation. Efficient state monitoring and anomaly detection may provide useful information for real-time decisions. With large amount of monitoring data, data-driven methods, especially deep learning methods are widely adopted for anomaly detection in various industrial applications. Although deep learning methods have advantages in discovering non-linear relations among complex and high-dimensional data, the large amount of hyperparameters can be hardly well tuned with a high computational burden for onboard computers. Therefore, in this work an efficient online anomaly detection model based on Broad Learning System (BLS) is established for detecting anomalies in the braking system. Furthermore, considering the intrinsic imbalanced data size on anomaly and normal states, the cost-sensitive learning method is integrated in the BLS model, for the first time. The proposed model is evaluated on real data collected from a high-speed train operating for one year, with respect to two performance metrics, i.e. G-mean and F1-score. Comparisons with benchmark neural networks and the combinations of sampling methods and BLS are also considered in this work.

show abstract

“…We did not test the effect of imbalanced classes in the training set and how it could affect the model's generalization capabilities. Note that imbalanced time series classification is a recent active area of research that merits an empirical study of its own [7]. At last, we should add that the number of generated time series in our framework was chosen to be equal to double the amount of time series in the most represented class (which is a hyper-parameter of our approach that we aim to further investigate in our future work).…”

Section: Data Augmentationmentioning

confidence: 99%

Data augmentation using synthetic data for time series classification with deep residual networks

Fawaz¹,

Forestier²,

Weber³

et al. 2018

Preprint

View full text Add to dashboard Cite

Data augmentation in deep neural networks is the process of generating artificial data in order to reduce the variance of the classifier with the goal to reduce the number of errors. This idea has been shown to improve deep neural network's generalization capabilities in many computer vision tasks such as image recognition and object localization. Apart from these applications, deep Convolutional Neural Networks (CNNs) have also recently gained popularity in the Time Series Classification (TSC) community. However, unlike in image recognition problems, data augmentation techniques have not yet been investigated thoroughly for the TSC task. This is surprising as the accuracy of deep learning models for TSC could potentially be improved, especially for small datasets that exhibit overfitting, when a data augmentation method is adopted. In this paper, we fill this gap by investigating the application of a recently proposed data augmentation technique based on the Dynamic Time Warping distance, for a deep learning model for TSC. To evaluate the potential of augmenting the training set, we performed extensive experiments using the UCR TSC benchmark. Our preliminary experiments reveal that data augmentation can drastically increase deep CNN's accuracy on some datasets and significantly improve the deep model's accuracy when the method is used in an ensemble approach.

show abstract

Cost-sensitive convolutional neural networks for imbalanced time series classification

Cited by 36 publications

References 46 publications

Deep learning for time series classification: a review

Deep learning for time series classification: a review

An Efficient Anomaly Detection for High-Speed Train Braking System Using Broad Learning System

Data augmentation using synthetic data for time series classification with deep residual networks

Contact Info

Product

Resources

About