Domain Transfer Multiple Kernel Boosting for Classification of EEG Motor Imagery Signals
TL;DR: A novel framework called domain transfer multiple kernel boosting (DTMKB), which extends the DTMKL algorithms by applying boosting techniques for learning kernel-based classifiers with the transfer of multiple kernels, which can be applied successfully in a small sample of EEG motor imagery signals.
Abstract: The application of wireless sensors in the brain-computer interface (BCI) system provides great convenience for the acquisition of electroencephalography (EEG) signals. However, a large amount of training data is needed to build the classification architectures used in motor imagery (MI) brain-computer interface (BCI), which is time-consuming to generate. To address this issue, transfer learning has gained significant attention in a small sample setting BCI system. The transfer learning methods have shown promising results by leveraging labeled patterns from the source domain to learn robust classifiers for the target domain, which has only a limited number of labeled samples. However, the successful application of such approaches in a motor imagery BCI remains limited. In this paper, we present a novel framework called domain transfer multiple kernel boosting (DTMKB), which extends the DTMKL algorithms by applying boosting techniques for learning kernel-based classifiers with the transfer of multiple kernels. Based on the proposed framework, we examined their empirical performance in comparison to several state-of-the-art algorithms on two MI task datasets. DTMKB yields the best performance for all datasets and achieves the best average classification accuracy 87.60%, 76.00%, 74.66%, and 74.13%, respectively. In particular, the proposed framework can be applied successfully in a small sample of EEG motor imagery signals.
Citations
More filters
TL;DR: Four main methods of transfer learning are described and their practical applications in EEG signal analysis in recent years are explored.
184 citations
TL;DR: The results show that TL can significantly improve the performance of decoding models across subjects/sessions and can reduce the calibration time of brain–computer interface (BCI) systems.
Abstract: The algorithms of electroencephalography (EEG) decoding are mainly based on machine learning in current research. One of the main assumptions of machine learning is that training and test data belong to the same feature space and are subject to the same probability distribution. However, this may be violated in EEG processing. Variations across sessions/subjects result in a deviation of the feature distribution of EEG signals in the same task, which reduces the accuracy of the decoding model for mental tasks. Recently, transfer learning (TL) has shown great potential in processing EEG signals across sessions/subjects. In this work, we reviewed 80 related published studies from 2010 to 2020 about TL application for EEG decoding. Herein, we report what kind of TL methods have been used (e.g., instance knowledge, feature representation knowledge, and model parameter knowledge), describe which types of EEG paradigms have been analyzed, and summarize the datasets that have been used to evaluate performance. Moreover, we discuss the state-of-the-art and future development of TL for EEG decoding. The results show that TL can significantly improve the performance of decoding models across subjects/sessions and can reduce the calibration time of brain–computer interface (BCI) systems. This review summarizes the current practical suggestions and performance outcomes in the hope that it will provide guidance and help for EEG research in the future.
36 citations
Cites methods from "Domain Transfer Multiple Kernel Boo..."
...MI [79] FTL Domain transfer multiple kernel boosting SVM BCIC-III-Iva 5 subjects’ data 81....
[...]
TL;DR: It is demonstrated that the proposed two-step single-trial classification method is effective to classify different movements on one arm, and provides the theoretic basis and technical support for the practical development of BMI-based motor restoration applications.
Abstract: Electroencephalography (EEG)-based brain-machine interface (BMI) is widely applied to control external devices like a wheel chair or a robotic arm, to restore motor function. EEG is useful to distinguish between left arm and right arm movements, however, it is difficult to classify the different movements on one arm. In this paper, a two-step single-trial classification method is proposed to recognize three movements (make a fist, hand extension and elbow flexion) of left and right arms: (1) distinguish between left arm and right arm movements by decoding event-related (de) synchronization (ERD/ERS) and (2) recognize the specific movement of this arm using corticomuscular coherence as features. Four healthy subjects are employed in a cue-based motor execution (ME) experiment. In Step one, ERD and post-movement ERS are found over the contralateral sensorimotor area; in Step two, for each movement, only the beta-band coherence between C3/C4 and the corresponding agonistic muscle is significant. The classification results show the best accuracy of Step one and Step two is 88.10% and 93.33%, respectively. This proposed method achieves a total accuracy of 82.22%. This study demonstrates that our method is effective to classify different movements on one arm, and provides the theoretic basis and technical support for the practical development of BMI-based motor restoration applications.
32 citations
TL;DR: An embedded lightweight SSVEP-BCI electric wheelchair with a hybrid hardware-driven visual stimulator is designed, which combines the advantages of liquid crystal display (LCD) and light-emitting diode (LED) to achieve lower energy consumption than the traditional LCD stimulator.
31 citations
TL;DR: This paper proves two novel spectral-similarity-based kernels based on spectral angle mapper (SAM) and spectral information divergence (SID) combined with the radial basis function (RBF) kernel to be Mercer’s kernels and considers three hyperspectral datasets to analyze their effectiveness.
Abstract: Spectral similarity measures can be regarded as potential metrics for kernel functions, and can be used to generate spectral-similarity-based kernels. However, spectral-similarity-based kernels have not received significant attention from researchers. In this paper, we propose two novel spectral-similarity-based kernels based on spectral angle mapper (SAM) and spectral information divergence (SID) combined with the radial basis function (RBF) kernel: Power spectral angle mapper RBF (Power-SAM-RBF) and normalized spectral information divergence-based RBF (Normalized-SID-RBF) kernels. First, we prove these spectral-similarity-based kernels to be Mercer’s kernels. Second, we analyze their efficiency in terms of local and global kernels. Finally, we consider three hyperspectral datasets to analyze the effectiveness of the proposed spectral-similarity-based kernels. Experimental results demonstrate that the Power-SAM-RBF and SAM-RBF kernels can obtain an impressive performance, particularly the Power-SAM-RBF kernel. For example, when the ratio of the training set is 20 % , the kappa coefficient of Power-SAM-RBF kernel (0.8561) is 1.61 % , 1.32 % , and 1.23 % higher than that of the RBF kernel on the Indian Pines, University of Pavia, and Salinas Valley datasets, respectively. We present three conclusions. First, the superiority of the Power-SAM-RBF kernel compared to other kernels is evident. Second, the Power-SAM-RBF kernel can provide an outstanding performance when the similarity between spectral signatures in the same hyperspectral dataset is either extremely high or extremely low. Third, the Power-SAM-RBF kernel provides even greater benefits compared to other commonly used kernels when the sizes of the training sets increase. In future work, multiple kernels combining with the spectral-similarity-based kernel are expected to be provide better hyperspectral classification.
26 citations
References
More filters
TL;DR: The relationship between transfer learning and other related machine learning techniques such as domain adaptation, multitask learning and sample selection bias, as well as covariate shift are discussed.
Abstract: A major assumption in many machine learning and data mining algorithms is that the training and future data must be in the same feature space and have the same distribution. However, in many real-world applications, this assumption may not hold. For example, we sometimes have a classification task in one domain of interest, but we only have sufficient training data in another domain of interest, where the latter data may be in a different feature space or follow a different data distribution. In such cases, knowledge transfer, if done successfully, would greatly improve the performance of learning by avoiding much expensive data-labeling efforts. In recent years, transfer learning has emerged as a new learning framework to address this problem. This survey focuses on categorizing and reviewing the current progress on transfer learning for classification, regression, and clustering problems. In this survey, we discuss the relationship between transfer learning and other related machine learning techniques such as domain adaptation, multitask learning and sample selection bias, as well as covariate shift. We also explore some potential future issues in transfer learning research.
18,616 citations
"Domain Transfer Multiple Kernel Boo..." refers background in this paper
...To address the above problems, transfer learning is a promising approach [8], [9]....
[...]
01 Aug 1997
TL;DR: The model studied can be interpreted as a broad, abstract extension of the well-studied on-line prediction model to a general decision-theoretic setting, and it is shown that the multiplicative weight-update Littlestone?Warmuth rule can be adapted to this model, yielding bounds that are slightly weaker in some cases, but applicable to a considerably more general class of learning problems.
Abstract: In the first part of the paper we consider the problem of dynamically apportioning resources among a set of options in a worst-case on-line framework. The model we study can be interpreted as a broad, abstract extension of the well-studied on-line prediction model to a general decision-theoretic setting. We show that the multiplicative weight-update Littlestone?Warmuth rule can be adapted to this model, yielding bounds that are slightly weaker in some cases, but applicable to a considerably more general class of learning problems. We show how the resulting learning algorithm can be applied to a variety of problems, including gambling, multiple-outcome prediction, repeated games, and prediction of points in Rn. In the second part of the paper we apply the multiplicative weight-update technique to derive a new boosting algorithm. This boosting algorithm does not require any prior knowledge about the performance of the weak learning algorithm. We also study generalizations of the new boosting algorithm to the problem of learning functions whose range, rather than being binary, is an arbitrary finite set or a bounded segment of the real line.
15,813 citations
TL;DR: Quantification of ERD/ERS in time and space is demonstrated on data from a number of movement experiments, whereby either the same or different locations on the scalp can display ERD and ERS simultaneously.
6,093 citations
"Domain Transfer Multiple Kernel Boo..." refers methods in this paper
...Thus, motor imagery has been widespread used as a major approach in BCI systems [3], [4]....
[...]
01 Dec 2000
TL;DR: It is demonstrated that spatial filters for multichannel EEG effectively extract discriminatory information from two populations of single-trial EEG, recorded during left- and right-hand movement imagery.
Abstract: The development of an electroencephalograph (EEG)-based brain-computer interface (BCI) requires rapid and reliable discrimination of EEG patterns, e.g., associated with imaginary movement. One-sided hand movement imagination results in EEG changes located at contra- and ipsilateral central areas. The authors demonstrate that spatial filters for multichannel EEG effectively extract discriminatory information from two populations of single-trial EEG, recorded during left- and right-hand movement imagery. The best classification results for three subjects are 90.8%, 92.7%, and 99.7%. The spatial filters are estimated from a set of data by the method of common spatial patterns and reflect the specific activation of cortical areas. The method performs a weighting of the electrodes according to their importance for the classification task. The high recognition rates and computational simplicity make it a promising method for an EEG-based brain-computer interface.
2,217 citations
"Domain Transfer Multiple Kernel Boo..." refers background in this paper
...Common spatial patterns (CSP) and spatial filters have the potential to achieve the rapid learning of proper training data, but do not perform well with a large amount of heterogeneous data recorded from other subjects or other sessions [17]....
[...]
10 Jul 2006
TL;DR: A novel statistical test of whether two samples are from the same distribution, compatible with both multivariate and structured data, that is fast, easy to implement, and works well, as confirmed by the experiments.
Abstract: Motivation: Many problems in data integration in bioinformatics can be posed as one common question: Are two sets of observations generated by the same distribution? We propose a kernel-based statistical test for this problem, based on the fact that two distributions are different if and only if there exists at least one function having different expectation on the two distributions. Consequently we use the maximum discrepancy between function means as the basis of a test statistic. The Maximum Mean Discrepancy (MMD) can take advantage of the kernel trick, which allows us to apply it not only to vectors, but strings, sequences, graphs, and other common structured data types arising in molecular biology. Results: We study the practical feasibility of an MMD-based test on three central data integration tasks: Testing cross-platform comparability of microarray data, cancer diagnosis, and data-content based schema matching for two different protein function classification schemas. In all of these experiments, including high-dimensional ones, MMD is very accurate in finding samples that were generated from the same distribution, and outperforms its best competitors. Conclusions: We have defined a novel statistical test of whether two samples are from the same distribution, compatible with both multivariate and structured data, that is fast, easy to implement, and works well, as confirmed by our experiments. Availability: Contact: [email protected]
1,315 citations
"Domain Transfer Multiple Kernel Boo..." refers methods in this paper
...[26] proposed an effective nonparametric criterion, referred to as the Maximum Mean Discrepancy (MMD), to compare data distributions based on the distance between the means of the samples from two domains in a kernel k induced reproducing kernel Hilbert space (RKHS) H , namely,...
[...]