Pregnancy outcome influence factor assessment method based on relative risk decision-making tree model

Abstract

The invention discloses a pregnancy outcome influence factor assessment method based on a relative risk decision-making tree model. The method comprises the steps that binary digital processing is conducted on data in a national free pre-pregnant eugenic health examination item information system, then a Pg exposure value of a pre-pregnant eugenic health examination-reproductive age population exposure value multi-dimensional input matrix is obtained through establishment, and a relative risk vector RR suitable for space-time multi-dimensional condition is established according to the Pg exposure value; a pre-pregnant eugenic health examination item Examy corresponding to the maximum relative risk in the RR is selected and serves as an empty father node of the relative risk decision-making tree model TR; blade node risk factor risk serves as an empty blade node of the relative risk decision-making tree model TR. The method is applied to pregnancy outcome influence factor assessment, the assessment accuracy of pregnancy outcome influence factors and their risk factors are effectively improved, the utilization value of pre-pregnant eugenic health examination data to smart city building is improved, and the method has the important significance on promotion of social harmonious and sustainable development.

Description

technical field [0001] The present invention relates to the technical field of pregnancy outcome, more particularly, to a method for evaluating factors affecting pregnancy outcome based on a relative risk decision tree model. Background technique [0002] "Algorithm Design Skills and Analysis" published in August 2004, translated by Wu Weichang, etc., on pages 209-211, discloses that "the usual expression of an algorithm consisting only of partitions is a binary tree called a decision tree". Decision Tree (DecisionTree) learning is a concept learning system CLS proposed by Hunt et al. in 1966 (ie Hunt E B, Marin J, Stone PJ. Experiments in induction. [J]. American Journal of Psychology, 1966, 80 (4): 17-19.) developed on the basis of learning the training set, the decision tree can mine useful rules and use them to predict new sets. It is a supervised, non-parametric machine learning method. Decision tree learning is one of the most widely used inductive reasoning methods, ...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problem that doctors are not clear about the factors affecting pregnancy outcome and their degree of influence in the risk assessment of pregnancy outcome, and it is difficult to comprehensively synthesize various information indicators, the present invention proposes a relative risk decision-making tree model based on pregnancy outcome. factor evaluation method

The method of the present invention solves the technical difficulties of evaluating data with multi-source heterogeneity and extremely high sparsity, and at the same time uses the current relative risk decision tree model to improve the selection of node attributes with information gain, which tends to select attributes with more values problem, so that the model can achieve better performance

Images