Data Mining Approach based on Neural Network and Decision Tree Methods for the Early Diagnosis of Risk of Gestational Diabetes Mellitus

Nowadays, in this industrial modern world, the incidence of chronic diseases has been significantly increased. Gestational diabetes mellitus is one of the major health problems that if not treated, it will cause serious complications for mother and her child. The purpose of this research was to find ways for determining the risk of gestational diabetes mellitus and making early diagnosis to prevent it in the initial stages of pregnancy.
This applied-survey research used two approaches of neural network and decision tree in experimental analysis of data and prediction. The extracted data were normalized and analyzed through Matlab software.
The results showed that data-based method is effective in improving the accuracy of prediction and has good performance in discovering implied knowledge and diagnosis of hidden relationships among data. In both methods, decision errors were acceptable and very close to each other.
Based on the obtained results, data mining methods can be used in health centers for less familiar diseases in order to achieve on-time diagnosis, patient management and to decrease treatment costs.