Reasearch on Pregnancy Hypertension Based on Data Mining

Pregnancy hypertension affects the safety of pregnant women and fetuses, and we apply data mining technology to facilitate model for various pregnancy indexes. We employ logistic regression, support vector machine and random forest to set up models for blood routine and biochemical indicators of 3000 pregnancy cases and evaluate the effectiveness of the models. Experimental results suggest that the accuracy of the support vector machine and random forest model are both 83% and that of the logistic regression is 81%, and the random forest model has best fitting precision. The results show that high body weight, edema and low calcium have higher connection with hypertension during pregnancy, suggesting the data mining is a promising method for gestational hypertension analysis.