Abstract:Realistic scenarios produce labeled data and unlabeled data, however, there are significant challenges in labeling time series data. It is imperative to effectively integrate the relationship between labeled and unlabeled data within semi-supervised classification model. This paper presents a novel semi-supervised classification method, namely Data Augmentation-Fast Shapelet Semi-Supervised Classification, which employs a data augmentation module to enhance the diversity of data and improve the generalization ability of the model, as well as a feature fusion module to enhance the semi-supervised network. A conditional generative adversarial network is used to synthesize excellent labeled time series samples to enhance the homogeneous data in the sample space, the fast shapelets method is used to quickly extract the important shape feature vectors in the time series, self-supervised and supervised learning are combined to fully learn the unlabeled and labeled data of the time series dataset. Thejoint loss function combines the loss functions of the two networks to optimize multiple objectives. Reinforcement learning is used to determine the weight coefficients of the joint loss function, at the same time, the reward function is modified to bias the supervisory loss, which improves the classification performance of the model under limited labeled data, and the model can also better achieve the semi-supervised classification task. The proposed method is validated on the UCR benchmark dataset, Electrocardiogram dataset, and Electroencephalogram dataset, the results show that the semi-supervised classification method can perform a more accurate semi-supervised classification of the time series, with an accuracy better than the comparison methods. Meanwhile, we use the plant electrical signal dataset obtained from actual measurements for testing, the visualizationanalysis can clearly show the model role in the semi-supervised classification task, and the experimental results fully demonstrate the effectiveness and applicability of the proposed method.

Active Learning for Multivariate Time Series Classification with Positive Unlabeled Data.

Time Series Feature Learning with Labeled and Unlabeled Data

Online Rule-Based Classifier Learning on Dynamic Unlabeled Multivariate Time Series Data

An Overview On Feature-Based Classification Algorithms For Multivariate Time Series

TimeAutoML: Autonomous Representation Learning for Multivariate Irregularly Sampled Time Series

Learning with Positive and Unlabeled Examples Using Topic-Sensitive PLSA.

Attention Based CNN-LSTM Network for Anomaly Pattern Classification of Multivariate Time Series

Unlabeled data selection for active learning in image classification

Pseudo Bidirectional Linear Discriminant Analysis For Multivariate Time Series Classification

Deep Reinforced Active Learning for Time Series Anomaly Detection.

TimeAutoAD: Autonomous Anomaly Detection With Self-Supervised Contrastive Loss for Multivariate Time Series

Semi-Supervised Text Classification Using Positive and Unlabeled Data

Heterogeneous Relationships of Subjects and Shapelets for Semi-supervised Multivariate Series Classification

Multi-instance Learning in the Presence of Positive and Unlabeled Bags

Self-labeling techniques for semi-supervised time series classification: an empirical study

Shapelet Based Two-Step Time Series Positive and Unlabeled Learning

Large-Margin Label-Calibrated Support Vector Machines for Positive and Unlabeled Learning

Effective Multi-Label Active Learning for Text Classification

Small Sample Time Series Classification Based on Data Augmentation and Semi-supervised Learning

Semi-unsupervised Learning for Time Series Classification

Unsupervised Classification of Multivariate Time Series Using VPCA and Fuzzy Clustering With Spatial Weighted Matrix Distance