People's way of life and diet have slowly changed as a result of the rapid growth of the economy in modern society. They have a bad habit and don't exercise enough, which has a negative impact on their health. In recent years, chronic diseases such as diabetes, cardiovascular sickness, and eternal kidney disease have become major public health issues worldwide. Huge published efforts in the medical field have indicated that diagnosis of these diseases at an early stage can greatly enhance patient outcomes and reduce healthcare costs. Furthermore, investigated efforts have also denoted that by consuming clinical data, data mining algorithms competently offer great potential in the prediction and diagnosis of a vast number of diseases. However, in a real-time frame, several approaches are available in an easy mode that are rapidly utilized by experts to predict and diagnose such diseases, but the associated flaws of former techniques such as some approaches taking a significant amount of time to compute while other methods are quick but inaccurate, have generate the need for additional research. This paper proposes a more precise clinical decision support model for the prediction of heart and diabetes disease with addressing the issue of existing methods. For this purpose, disease affected patient data has accessed from UCI's open repositories and taking into account the disease prediction capabilities of five classifiers, including NB, DT, SVM, IBK, and ANN. Additionally, the standard K-fold cross-validation method has used to train and validate the proposed processes. Attained upsurges amount of disease prediction accuracy have symbolized the proficiency of the proposed approach with attaining accuracy 94% for recognize diabetic patients and 93% accuracy for identify heart illness.