Analysis of application of machine learning based on bibliometrics in pattern recognition

Abstract

This paper analyzes the application of machine learning in pattern recognition by using bibliometrics methods, keyword cluster analysis, and burst terms detection. According to analysis, the number of published studies on the application of machine learning in pattern recognition has shown a rapid growth trend in recent years. The top five countries with published volumes are the United States, China, the United Kingdom, Germany, and India. According to the results of keyword co-occurrence cluster analysis, it is found that the hot research topics in this field mainly include feature extraction, support vector machines, and neural networks. Research topics include functional magnetic resonance imaging, deep learning and image processing, bioinformatics and fuzzy logic systems, semi-supervised learning and face recognition. According to the analysis of burst terms detection results, it is found that future research frontiers include big data, sensors, classifier integration, entropy, pattern classification, attribute reduction, image analysis, principal component analysis, fuzzy logic, Alzheimer’s disease, magnetic resonance imaging, etc.