Comparing classification algorithms of data mining in diagnosis of diabetes and assessing the effectiveness of k-fold cross validation in the accuracy of the constructed model

Comparing classification algorithms of data mining in diagnosis of diabetes and assessing the effectiveness of k-fold cross validation in the accuracy of the constructed model

One of the applications of data mining is in medicine and model construction for disease diagnosis. The more the model learns from previous data, the more accurate it would perform. The essential issue is that, the training and testing data in classification of data must be selected in a way that the model enjoys the most efficient learning from previous data and the highest accuracy in diagnosis of the disease. In this study, the Pima dataset of diabetics is applied, the models for predicting and diagnosing diabetes are developed based on KNN, SVM, Nave Bayesian and Decision Tree classification methods and the accuracy of each model is evaluated. The effectiveness of k-fold validation on the accuracy of each model is assessed. According to the findings here, k-fold cross validation increases the model accuracy and a classification technique would not always have the best performance and accuracy, while it depends on the nature and complexity of the dataset. The simulation is made by the tool named RapidMiner.

One of the applications of data mining is in medicine and model construction for disease diagnosis. The more the model learns from previous data, the more accurate it would perform. The essential issue is that, the training and testing data in classification of data must be selected in a way that the model enjoys the most efficient learning from previous data and the highest accuracy in diagnosis of the disease. In this study, the Pima dataset of diabetics is applied, the models for predicting and diagnosing diabetes are developed based on KNN, SVM, Nave Bayesian and Decision Tree classification methods and the accuracy of each model is evaluated. The effectiveness of k-fold validation on the accuracy of each model is assessed. According to the findings here, k-fold cross validation increases the model accuracy and a classification technique would not always have the best performance and accuracy, while it depends on the nature and complexity of the dataset. The simulation is made by the tool named RapidMiner.