Diabetes Recognition in Pregnant Women by Extracting Features Using PCA and Data Mining Algorithms

In this world diabetes is the most serious immedicable disease. This disease oppressed a huge amount of pregnant woman in the world day by day. The blood sugar levels is also affected for this disease. For this reason the disease will dispatched to their unborn baby. For this purpose the research of diabetes instances have been confined to pregnant women in this paper. To prophesy diabetes by using various data mining algorithms PIMA INDIAN DIABETES dataset has become a standard to gain better result by those algorithms. In this work, five data mining algorithm has been used where support vector machine algorithm gives higher accuracy compared to others. After finding feature importance value for each of the attribute principal component analysis technique is used for feature selection. Those we obtained the highest accuracy and fl-score in SVM which is 87% amongst all the techniques.