Penerapan Backpropagation Jaringan Saraf Tiruan untuk Prediksi Diabetes menggunakan dataset Pima Indians

Abstract

The application of Backpropagation Neural Network (BPNN) for diabetes prediction using the Pima Indians dataset is a crucial step towards developing effective diagnostic tools. This study preprocesses the dataset by addressing missing values and aoutliers through statistical techniques and scalling. The data is then split into training and testing sets , followed by feature scalling using StandartScaller. A grid search with cross-validation is performed to optimize the hyperparameters of the Multilayer Perceptron (MLP) classifier. The optimized MLP classifier achieved a best cross-validation score of 78.50%. on the test set, the classifier obtained an accuracy of 83%, with a precision of 0.85 for the non-diabetes class and 0.79 for the diabetes class. The performance is evaluated using classification reports and confusion matrices, highliting the potential of BPNN in medical diagnosis. The findings contribute to the ongoing research in applying machine learning to healthcare. Emphasizing the importance of data preprocessing and hyperparameter tuning.

Keywords: Backpropagation , Diabetes Prediction, Pima Indians Dataset , Neural Network , Hyperparameter Tuning , Machine Learning, Healthcare