Collaborative Filtering Algorithm based on Hybrid Machine Learning Optimization

Abstract

Under the influence of the mobile Internet era, users' patience is increasingly limited. Future recommendation algorithms should quickly respond to users' urgent needs to save users' time. Under the background of big data, how to ensure the relatively low complexity and high accuracy of information push and display is a very valuable topic. Therefore, this paper discusses and analyzes the collaborative filtering algorithm (CFA) based on hybrid machine learning (ML) optimization. In this paper, the research background and significance of CFA are first described, including the development status of collaborative filtering recommendation algorithm, the research status of association rule recommendation algorithm and particle swarm optimization algorithm, and the problems in collaborative filtering recommendation algorithm. A new hybrid CFA model is proposed based on the hybrid ML optimization. This paper designs a simulation table of three recommendation algorithms to verify the proposed CFA. The experiment shows that the recall rate of the CFA proposed in this paper based on the hybrid ML optimization is not lower than other recommendation algorithms, which verifies the effectiveness of this algorithm.