Early detection of diabetes patients using soft computing

In this paper, diagnosis of diabetes using soft computing is presented. This research work is based on the fuzzy if-then rules and tuning of the parameters by neural network. A variational method for determining, globally optimal learning parameters and learning rules for on-line gradient descent training has been proposed in the paper. Neurofuzzy system is put in the framework to facilitate learning and adaptation for reducing the error in the output. A knowledge based system has been developed in client server for analysis of the disease and for storing the corresponding solution into the database. Simulated results show the proposed work is effective and after the analysis of the diagnosis result of the patients, the client server sends a message for first aid treatment.