An Effective Approach for Detecting Diabetes using Deep Learning Techniques based on Convolutional LSTM Networks

Abstract

The most common disorder affecting millions of population worldwide due to insufficient release of insulin by pancreas is diabetes. Early detection or precaution of diabetes is necessary, otherwise leads to many complicated problems. Predicting diabetes at early stages with appropriate treatment, individuals can maintain a happy life. If the conventional diabetes detection method is tedious, the identification of diabetes from clinical and physical data requires an automated system. This paper proposes an approach to enhance diabetes prediction using deep learning techniques. Based on the Convolutional Long Short-term Memory (CLSTM), we developed a diabetes classification model and compared with the existing methods on the Pima Indians Diabetes Database (PIDD). We assessed the findings of various classification approaches in this study. The proposed approach is further improved by an efficient pre-processing mechanism called multivariate imputation by chained equations. The outcomes are promising compared to existing machine learning approaches and other research models.