[1K4-E-1] Knowledge engineering

[1K4-E-1-02] Multimodal Neural Network–based Health Platform for Knowledge Decision-Making

〇Kyungyong Chung¹ (1. Kyonggi University)

キーワード：Multimodal Neural Network, Health Platform, Data Mining

There is a need for artificial intelligence-oriented information technologies (aimed at continuous monitoring and life-care of chronic diseases through health platforms) that can discover potential health-risk-factor changes and predict emerging risks. In this paper, we propose a multimodal neural network-based health platform for knowledge decision-making. The proposed method learns the relationships present between heterogeneous data and the multimodal neural network, and extracts the common information shared between the modals to estimate health-risk factors. The correlation of variables appearing in the health platform is used to construct a multimodal neural network, and shared common information is combined to estimate the health-risk factors. The correlations of the variables are shown as positive correlations and negative correlations. A positive correlation indicates a relationship in which two variables change in the same direction, and a negative correlation indicates a relationship in which they change in a different direction. The proposed multimodal neural network is used to solve the health-risk–factor problem in the health platform, improving the reliability of the data.