Novel diabetes classification approach based on CNN-LSTM: enhanced performance and accuracy

This paper deals with the development of an approach for diabetes classification harnessing ConvolutionalNeural-network (CNN) and a Long-Short-Term-Memory (LSTM) model. The proposed method harnesses the strengths of LSTM and CNN architectures to effectively capture sequential patterns and extract meaningful features from the input data. A comprehensive dataset containing relevant features for diabetes patients is used to train and evaluate the classifiers. Evaluation metrics such as kappa score, F1-score, accuracy, precision, and recall are employed in ordre to assess the performance of each model. The results demonstrate that the CNNLSTM model outperforms other models, including Logistic Regression, Random Forest, SVM, and KNN, achieving an impressive accuracy of 97%. These findings shed light on the effectiveness of the proposed approach in accurately classifying diabetes, resulting in significant advancement in diabetes diagnosis and treatment and opening up exciting possibilities for personalized healthcare.