Mental Health Education Model Based on Matrix Decomposition

Abstract

The exploration on whether it is possible to use the psychological trust relation in the
rough open psychological system when the conditions for the use of factor analysis and
the multiple regression method are not satisfied, is fundamentalally one of the most
complicated social-psychological relations. It involves many factors such as hypothesis,
expectation, behavior, and environment. Hence, accurate quantified representation and
prediction are tricky. Combined with the cognitive behavior in the human society, a
matrix decomposition mental health education model that conforms to the human
cognitive habit is proposed: (1) Through the historical evidence window-based adaptive
credibility decision, not only the common subjective judgment of weight in the present
models can be overcome, but the reliability prediction issue when the direct evidence is
insufficient can be resolved; (2) The DTT (direct trust tree) mechanism is used to search
and aggregate the global psychological feedback information, to reduce the consumption
of the negative psychological factors, enhance the scalability of the psychological
system in the large scale psychological system; (3) We introduced the concept of the
induced ordered average weighted (abbreviated as IOWA) operator and established an
IOWA operator-based psychological trust prediction model to address the poor dynamic
adaptability of conventional prediction models. From the results of tests, it can be seen
that the model proposed in this paper has greater dynamic adaptability and higher
prediction precision than the existing models.