Collaborative Filtering Algorithms in Financial Performance Evaluation Indicators for Intelligent Innovation Dynamic Research

Abstract

Abstract This study develops a financial performance evaluation system for S Group, employing a collaborative filtering algorithm to address the limitations inherent in traditional financial performance evaluation methodologies. Improvements have been made to the similarity measure method, data imputation technique, and rating prediction approach within the model. Subsequently, the financial status of S Group is scrutinized, with a thorough exploration of the input and output indicators, leading to the calculation of the efficiency values within the financial indicator system. Through sensitivity analysis, this research investigates the impact of input and output indicators on the outcomes derived from collaborative filtering. Based on the study, recommendations are formulated in alignment with the evaluation outcomes. The findings reveal that the average comprehensive efficiency of S Group over the past decade is 0.772, with a median comprehensive efficiency of 0.788, indicative of a robust financial condition. This research provides a valuable reference for the formulation of financial performance evaluation systems and the analysis of financial performance across various industries.