Development of R-Shiny interface for implementation of backpropagation neural network model in breast cancer classification

Abstract

Abstract Artificial Neural Network or Neural Network (NN) is an information processing system that has similar characteristics to the neural network in living things. One type of NN that is often used in classification is Backpropagation Neural Network (BPNN). BPNN is an NN model that is often used for classification because it does not need to use assumptions and has high accuracy. One of the classification problems that can be solved with BPNN is the classification of breast cancer. The breast cancer data used in this study came from the UCI Machine Learning website. The problem with BPNN is that programming is difficult for users who do not understand the program, especially the R program. Therefore, to make it easier for users to analyze BPNN, an R-Shiny application or interface is created using the RStudio program. The application or R-Shiny interface that has been created has several advantages, namely the application process that is fast in displaying classification results, the use of user-friendly applications and the use of applications that are more comfortable when compared to having to write syntax such as in the R program. BPNN classification results use The R-Shiny interface has a different level of accuracy for each experiment due to the random distribution of training & testing data. The experiments conducted in this study resulted in a range of accuracy values ranging from 58.33% to 91.67% with an average accuracy of 74.17%.