Disease diagnosis using meta-learning framework

Data mining techniques have been extensively used in medical decision support systems for the purpose of prediction and to correctly identify severaldiseases correctly. These techniques are applicable for designing health related systems because of their capability to find outthe concealed patterns and associations in health related data. The main objective of this paper is to develop and implement a framework which provides considerable classification results for users who have no prior data mining knowledge. We also propose a proper prediction model to improve the reliability of medical assessment and medications for ailments. We analyzed different medical records for certain disease and based on the hypothesis made on the training dataset, applied it on the test dataset and achieved disease with a good accuracy. We focus on reducing the dependency of the system on user input, and offer the capability of a guided search for a proper learning algorithm through performance metrics.