Self-recommendation System of Endowment Pattern Based on Neural Network for China

摘要

With the urbanization acceleration and the development of the endowment industry, people's understanding of the endowment pattern is getting deeper and deeper. Therefore, people are no longer satisfied with the simple and traditional family endowment pattern, but prefer some new endowment patterns to meet the needs of people's diversified endowment. This paper makes a thorough and in-depth study on the existing endowment patterns in China, and sums up ten typical endowment patterns with their own characteristics. In this paper, we did a large sample survey on the ten typical endowment patterns, and recover 1531 valid questionnaires. We obtained the important evidence about how people choose the endowment pattern based on analysis of the effective recovery questionnaires. In this paper, we built up an endowment pattern prediction model based on BP feed-forward neural network. By calculation and analysis, an 87.8% prediction accuracy self-recommendation system of endowment patterns successfully constructed. The self-recommendation system can well predict people's choice of endowment patterns, which is of great significance to judge the future development of the endowment industry.