ChronoNet: A Deep Recurrent Neural Network for Abnormal EEG Identification

Roy, Subhrajit; Kiral-Kornek, Isabell; Harrer, Stefan

doi:10.1007/978-3-030-21642-9_8

Cited by 133 publications

(119 citation statements)

References 26 publications

Supporting

Mentioning

114

Contrasting

Unclassified

Order By: Relevance

“…To the best of our knowledge, our work is the first attempt to study few-shot learning for TSC. We formulate the few-shot learning problem for UTSC, and build on top of the following recent advances in deep learning research to develop an effective few-shot approach for TSC: i) gradient-based meta-learning [11,24], ii) residual network with convolutional layers for TSC [40], iii) leveraging multi-length filters to ensure generalizability of filters to tasks with varying time series length and temporal properties [18,29], and iv) triplet loss [35] to ensure generalizability to tasks with varying number of classes without introducing any additional parameters.…”

Section: Related Workmentioning

confidence: 99%

“…In order to quickly adapt to any unseen task, the neural network should be able to extract temporal features at multiple time scales and should ensure that the fine-tuned network can generalize to time series of varying lengths across tasks. We, therefore, use filters of multiple lengths in each convolutional block to capture temporal features at various time scales, as found to be useful in [3,18,29].…”

Section: Neural Networkmentioning

confidence: 99%

See 1 more Smart Citation

Meta-Learning for Few-Shot Time Series Classification

Narwariya¹,

Malhotra²,

Vig³

et al. 2020

Proceedings of the 7th ACM IKDD CoDS and 25th COMAD

View full text Add to dashboard Cite

Deep neural networks (DNNs) have achieved state-of-the-art results on time series classification (TSC) tasks. In this work, we focus on leveraging DNNs in the often-encountered practical scenario where access to labeled training data is difficult, and where DNNs would be prone to overfitting. We leverage recent advancements in gradientbased meta-learning, and propose an approach to train a residual neural network with convolutional layers as a meta-learning agent for few-shot TSC. The network is trained on a diverse set of fewshot tasks sampled from various domains (e.g. healthcare, activity recognition, etc.) such that it can solve a target task from another domain using only a small number of training samples from the target task. Most existing meta-learning approaches are limited in practice as they assume a fixed number of target classes across tasks. We overcome this limitation in order to train a common agent across domains with each domain having different number of target classes, we utilize a triplet-loss based learning procedure that does not require any constraints to be enforced on the number of classes for the few-shot TSC tasks. To the best of our knowledge, we are the first to use meta-learning based pre-training for TSC. Our approach sets a new benchmark for few-shot TSC, outperforming several strong baselines on few-shot tasks sampled from 41 datasets in UCR TSC Archive. We observe that pre-training under the meta-learning paradigm allows the network to quickly adapt to new unseen tasks with small number of labeled instances.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Neural Networkmentioning

confidence: 99%

Meta-Learning for Few-Shot Time Series Classification

Narwariya¹,

Malhotra²,

Vig³

et al. 2020

Proceedings of the 7th ACM IKDD CoDS and 25th COMAD

View full text Add to dashboard Cite

show abstract

“…Generally, the CNN was applied for better classification performances to various types of images. Recently, the CNN architecture was applied into the BMI fields to consider dynamics of the signal during the movement and to extract static energy feature robustly [32].…”

Section: E Data Analysismentioning

confidence: 99%

Classification of Upper Limb Movements Using Convolutional Neural Network with 3D Inception Block

Lee

Jeong

Shim

et al. 2020

2020 8th International Winter Conference on Brain-Computer Interface (BCI)

View full text Add to dashboard Cite

A brain-machine interface (BMI) based on electroencephalography (EEG) can overcome the movement deficits for patients and real-world applications for healthy people. Ideally, the BMI system detects user movement intentions transforms them into a control signal for a robotic arm movement. In this study, we made progress toward user intention decoding and successfully classified six different reaching movements of the right arm in the movement execution (ME). Notably, we designed an experimental environment using robotic arm movement and proposed a convolutional neural network architecture (CNN) with inception block for robust classify executed movements of the same limb. As a result, we confirmed the classification accuracies of six different directions show 0.45 for the executed session. The results proved that the proposed architecture has approximately 6∼13% performance increase compared to its conventional classification models. Hence, we demonstrate the 3D inception CNN architecture to contribute to the continuous decoding of ME.

show abstract

“…In this work, we propose ConvTimeNet (CTN), a deep CNNbased transfer learning approach for UTSC. CTN consists of multiple length 1-D convolutional filters in all convolutional layers (similar to that in InceptionNet [13], [14]) resulting in filters that can capture features at multiple time scales. The key contributions of this work can be summarized as follows:…”

Section: Introductionmentioning

confidence: 99%

ConvTimeNet: A Pre-trained Deep Convolutional Neural Network for Time Series Classification

Kashiparekh

Narwariya²,

Malhotra³

et al. 2019

2019 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

Training deep neural networks often requires careful hyper-parameter tuning and significant computational resources. In this paper, we propose ConvTimeNet (CTN): an offthe-shelf deep convolutional neural network (CNN) trained on diverse univariate time series classification (TSC) source tasks. Once trained, CTN can be easily adapted to new TSC target tasks via a small amount of fine-tuning using labeled instances from the target tasks. We note that the length of convolutional filters is a key aspect when building a pre-trained model that can generalize to time series of different lengths across datasets. To achieve this, we incorporate filters of multiple lengths in all convolutional layers of CTN to capture temporal features at multiple time scales. We consider all 65 datasets with time series of lengths up to 512 points from the UCR TSC Benchmark for training and testing transferability of CTN: We train CTN on a randomly chosen subset of 24 datasets using a multi-head approach with a different softmax layer for each training dataset, and study generalizability and transferability of the learned filters on the remaining 41 TSC datasets. We observe significant gains in classification accuracy as well as computational efficiency when using pre-trained CTN as a starting point for subsequent task-specific fine-tuning compared to existing state-of-the-art TSC approaches. We also provide qualitative insights into the working of CTN by: i) analyzing the activations and filters of first convolution layer suggesting the filters in CTN are generically useful, ii) analyzing the impact of the design decision to incorporate multiple length decisions, and iii) finding regions of time series that affect the final classification decision via occlusion sensitivity analysis.

show abstract

ChronoNet: A Deep Recurrent Neural Network for Abnormal EEG Identification

Cited by 133 publications

References 26 publications

Meta-Learning for Few-Shot Time Series Classification

Meta-Learning for Few-Shot Time Series Classification

Classification of Upper Limb Movements Using Convolutional Neural Network with 3D Inception Block

ConvTimeNet: A Pre-trained Deep Convolutional Neural Network for Time Series Classification

Contact Info

Product

Resources

About