A Community Based Study for Early Detection of Postpartum Depression using Improved Data Mining Techniques

Abstract

Pregnancy for women is one of the most beautiful feelings that exist in world. But this pregnancy leads women to various hormonal, physical and mental changes which affect their life, family, child and many more. Immediate after delivery the women had to overcome the rapid slowdown of hormones and initial Postpartum Blues. Today it has been observed that Postpartum Blue when exist for more month and year are predict to be suffering from Postpartum Depression or Postpartum Psychosis. The study tried to generate the most possible condition on which the women will suffer from Postpartum Depression by taking Survey of 96 participants. The study tried to develop a predictive model which can help to predict the Postpartum Depression among women. The Predictive model development is done using Data Mining Algorithms-J48, Random Tree, Random Forest and Reduce Error Pruning (REP) Tree. These four algorithms are further collaborated with Adaptive Boosting and Bagging. The development of class model in dataset is done by Edinburgh Postpartum Depression Scale which help to justify the exact observation of suffering from Postpartum Depression. The development of model is done using WEKA application tool.