Epilepsy is a neurological illness caused by abnormal discharge of brain neurons, where epileptic seizure can lead to life-threatening emergencies. By analyzing the encephalogram (EEG) signals of patients with epilepsy, their conditions can be monitored and seizure can be detected and intervened in time. As the identification of effective features in EEG signals is important for accurate seizure detection, this paper proposes a multi-view deep feature extraction method in attempt to achieve this goal. The method first uses fast Fourier transform (FFT) and wavelet packet decomposition (WPD) to construct the initial multi-view features. Convolutional neural network (CNN) is then used to automatically learn deep features from the initial multi-view features, which reduces the dimensionality and obtain the features with better seizure identification ability. Furthermore, the multi- view Takagi-Sugeno-Kang fuzzy system (MV-TSK-FS), an interpretable rule-based classifier, is used to construct a classification model with strong generalizability based on the deep multi- view features obtained. Experimental studies show that the classification accuracy of the proposed multi-view deep feature extraction method is at least 1% higher than that of common feature extraction methods such as principal component analysis (PCA), FFT and WPD. The classification accuracy is also at least 4% higher than the average accuracy achieved with single-view deep features.

In many practical applications of classifiers, not only high accuracy but also high interpretability is required. Among a wide variety of existing classifiers, Takagi–Sugeno–Kang (TSK) fuzzy classifiers may be one of the best choices for achieving a good balance between interpretability and accuracy. In order to further improve their accuracy without losing their interpretability, we propose a highly interpretable deep TSK fuzzy classifier HID-TSK-FC (deep shared-linguistic-rule-based TSK fuzzy classifier) based on the concept of shared linguistic fuzzy rules. The proposed classifier has two characteristics: One is a stacked hierarchical structure of component TSK fuzzy classifiers for high accuracy, and the other is the use of interpretable linguistic rules with the same set of linguistic labels for all inputs. High interpretability is achieved at each layer by using the same set of linguistic values for all inputs, including the outputs from the previous layers in the stacked hierarchical structure. We show that a linguistic rule with the outputs from the previous layers as its inputs is equivalent to a fuzzy rule with a nonlinear consequent or a linear consequent with a certainty factor. We also show that HID-TSK-FC is mathematically equivalent to a novel TSK fuzzy classifier with shared interpretable linguistic fuzzy rules. Promising performance of HID-TSK-FC is demonstrated through extensive computational experiments on benchmark datasets and a real-world application case.

Although Takagi-Sugeno-Kang (TSK) fuzzy classifier has been applied to a wide range of practical scenarios, how to enhance its classification accuracy and interpretability simultaneously is still a challenging task. In this paper, based on the powerful stacked generalization principle, a deep TSK fuzzy classifier (D-TSK-FC) is proposed to achieve the enhanced classification accuracy and triplely concise interpretability for fuzzy rules. D-TSK-FC consists of base-building units. Just like the existing popular deep learning, D-TSK-FC can be built in a layer-by-layer way. In terms of the stacked generalization principle, the training set plus random shifts obtained from random projections of prediction results of current base-building unit are presented as the input of the next base-building unit. The hidden layer in each base-building unit of D-TSK-FC is represented by triplely concise interpretable fuzzy rules in the sense of randomly selected features with the fixed five fuzzy partitions, random rule combinations, and the same input space kept in every base-building unit of D-TSK-FC. The output layer of each base-building unit can be learnt quickly by least learning machine (LLM). Besides, benefiting from LLM, D-TSK-FC's deep learning can be well scaled up for large datasets. Our extensive experimental results witness the power of the proposed deep TSK fuzzy classifier.

Intelligent recognition of electroencephalogram (EEG) signals is an important means to detect seizure. Traditional methods for recognizing epileptic EEG signals are usually based on two assumptions: 1) adequate training examples are available for model training, and 2) the training set and the test set are sampled from datasets with the same distribution. Since seizures occur sporadically, training examples of seizures could be limited. Besides, the training and test sets are usually not sampled from the same distribution for generic non-patient-specific recognition of EEG signals. Hence, the two assumptions in traditional recognition methods could hardly be satisfied in practice, which results in degradation of model performance. Transfer learning is a feasible approach to tackle this issue attributed to its ability to effectively learn the knowledge from the related scenes (source domains) for model training in the current scene (target domain). Among the existing transfer learning methods for epileptic EEG recognition, transductive transfer learning fuzzy systems (TTL-FSs) exhibit distinctive advantages – the interpretability that is important for medical diagnosis and the transfer learning ability that is absent from traditional fuzzy systems. Nevertheless, the transfer learning ability of TTL-FSs is restricted to a certain extent since only the discrepancy in marginal distribution between the training data and test data is considered. In this study, the enhanced transductive transfer learning Takagi-Sugeno-Kang fuzzy system construction method (ETTL-TSK-FS) is proposed to overcome the challenge by introducing two novel transfer learning mechanisms: (1) joint-knowledge is adopted to reduce the discrepancy between the two domains, and (2) an iterative transfer learning procedure is introduced to enhance transfer learning ability. Extensive experiments have been carried out to evaluate the effectiveness of the proposed method in recognizing epileptic EEG signals on the Bonn and CHB-MIT EEG datasets. The results show that the method is superior to or at least competitive with some existing state-of-art methods under the scenario of transfer learning.

Autism spectrum disorder (ASD) is an ageand sex-related neurodevelopmental disorder that alters the brain's functional connectivity (FC). The changes caused by ASD are associated with different age and sex-related patterns in neuroimaging data. However, most contemporary computer-assisted ASD diagnosis methods ignore the aforementioned age-/sexrelated patterns. In this paper, we propose a novel sparse multiview task-centralized (Sparse-MVTC) ensemble classification method for image-based ASD diagnosis. Specifically, with the age and sex information of each subject, we formulate the classification as a multitask learning problem, where each task corresponds to learning upon a specific age/sex group. We also extract multiview features per subject to better reveal the FC changes. Then, in Sparse-MVTC learning, we select a certain central task and treat the rest as auxiliary tasks. By considering both task-task and view-view relationships between the central task and each auxiliary task, we can learn better upon the entire dataset. Finally, by selecting the central task, in turn, we are able to derive multiple classifiers for each task/group. An ensemble strategy is further adopted, such that the final diagnosis can be integrated for each subject. Our comprehensive experiments on the ABIDE database demonstrate that our proposed Sparse-MVTC ensemble learning can significantly outperform the state-of-the-art classification methods for ASD diagnosis.

Due to its strong performance in handling uncertain and ambiguous data, the fuzzy k-nearest-neighbor method (FKNN) has realized substantial success in a wide variety of applications. However, its classification performance would be heavily deteriorated if the number k of nearest neighbors was unsuitably fixed for each testing sample. This study examines the feasibility of using only one fixed k value for FKNN on each testing sample. A novel FKNN-based classification method, namely, fuzzy KNN method with adaptive nearest neighbors (A-FKNN), is devised for learning a distinct optimal k value for each testing sample. In the training stage, after applying a sparse representation method on all training samples for reconstruction, A-FKNN learns the optimal k value for each training sample and builds a decision tree (namely, A-FKNN tree) from all training samples with new labels (the learned optimal k values instead of the original labels), in which each leaf node stores the corresponding optimal k value. In the testing stage, A-FKNN identifies the optimal k value for each testing sample by searching the A-FKNN tree and runs FKNN with the optimal k value for each testing sample. Moreover, a fast version of A-FKNN, namely, FA-FKNN, is designed by building the FA-FKNN decision tree, which stores the optimal k value with only a subset of training samples in each leaf node. Experimental results on 32 UCI datasets demonstrate that both A-FKNN and FA-FKNN outperform the compared methods in terms of classification accuracy, and FA-FKNN has a shorter running time.