Machine learning for data fusion

• Main Authors: Gupta, Anurag, Gupta, Rupal, Saxena, Vineet
• Format: Conference Proceeding
• Language: English
• Subjects: Data integration; Machine learning; Synthesis
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0125054

Fusion of data is a common way of dealing with incomplete raw data to collect reliable, useful, and accurate data. Compared with a variety of the traditional stochastic fusion techniques, a machine-learning approach which automatically learns without explicit programming from past experiences, remarkably refreshed the process of fusion by providing efficient calculation and prediction. However, a comprehensive analysis of the current developments in machine knowledge for information synthesis still lacks in the literature. Reviewing and summarizing the state of the art is also useful for having an unfathomable vision into How to support and enhance information synthesis in machine learning. This paper provides an overview of data fusion approaches in depth for machine learning. Conceived, introduced, architecture, scientific and conventional methods. This paper firstly gives a comprehensive outline to the context of machine learning and information fusion. Then suggest a range of specifications and use them as guidelines for testing and assessing the output based on machine learning of existing fusion methods.