Research on Machine Learning Based on Multi-label Algorithm
Kexuan Wang, Xin Yifei, Xinyu Xiong · 2020
In recent years, multi-label learning has performed well in solving related problems such as a single object may have multiple associated categories, which has attracted widespread attention from researchers. This paper applies data mining to the information and needs of different users. Analyze the correlation between many label features, and use the sample clustering information to adjust the similarity matrix of weakly labelled samples. This paper proposes a measure for classifier credibility, and constructs a similarity based on k-means. Matrix, and introduce different levels of multi-example learning algorithms as our classifier model, and propose two classification distances: the minimum classification distance of labels and the average classification distance of labels. Finally, we apply the model to natural scene image classification and text classification. After comparing the method proposed in this paper with the ordinary sample, it is found that the method in this paper can obtain better classification performance in various evaluation indicators under the same training sample.