Online Multi-Label Streaming Feature Selection Based on Label Group Correlation and Feature Interaction

Abstract

Multi-label streaming feature selection has received widespread attention in recent years because the dynamic acquisition of features is more in line with the needs of practical application scenarios. Most previous methods either assume that the labels are independent of each other, or, although label correlation is explored, the relationship between related labels and features is difficult to understand or specify. In real applications, both situations may occur where the labels are correlated and the features may belong specifically to some labels. Moreover, these methods treat features individually without considering the interaction between features. Based on this, we present a novel online streaming feature selection method based on label group correlation and feature interaction (OSLGC). In our design, we first divide labels into multiple groups with the help of graph theory. Then, we integrate label weight and mutual information to accurately quantify the relationships between features under different label groups. Subsequently, a novel feature selection framework using sliding windows is designed, including online feature relevance analysis and online feature interaction analysis. Experiments on ten datasets show that the proposed method outperforms some mature MFS algorithms in terms of predictive performance, statistical analysis, stability analysis, and ablation experiments.

Research on Three-Phase Asynchronous Motor Fault Diagnosis Based on Multiscale Weibull Dispersion Entropy

“…Entropy is very sensitive to the nonlinear features of a signal, and traditional time-frequency domain features may not be able to capture these nonlinear features efficiently [6]. Entropy is relatively robust to noise (traditional time-frequency domain features are susceptible to noise) and is able to resist the influence of noise to a certain extent and more accurately reflect the characteristics of a signal [7]. Steven M. Pincus proposed approximate entropy [8] from the perspective of measuring the complexity of a signal sequence as a measure of the magnitude of the probability of generating a new pattern in the signal; however, this approach involves a comparison with its own vectors, which is incompatible with the new information viewpoint and can be biased.…”