Prediksi Pendapatan Penjualan Obat Menggunakan Metode Backpropagation Neural Network dengan Algoritma Genetika Sebagai Seleksi Fitur

  • Nur Azise
  • Pulung Nurtantio Andono
  • Ricardus Anggi Pramunendar
Keywords: drug sales revenue predictions, artificial neural Network, genetic algorithm, bacpropagation, selection of features


The hospital is one of the means of health services for the community, in which there are multiple units, one of which was the installation of a pharmaceutical is a source of revenue for the hospitals contributed by 40 – 60%. Each month the sale of drugs on pharmaceutical erratic installation (fluctuating) and have an impact on earning spharma ceutical installations specifically and at hospitals in General, i.e. against the determination of the lead in policy development and the development of hospitals inthe future. Therefore the forecast or prediction about the drug's sales revenue is urgently needed. Forecasting technique commonly used is the technique of forecastingwith Artificial neural Network method or so-called Artificial neural Network which hasthe best accuracy with the value error. However, the method of Artificial neural Network has a number of shortcomings, so it takes an optimization method, one of themwith Genetic Algorithm optimization methods. In this study using data on drug sales revenue installation hospital Elizabeth Situbondo. In the process of training and testing data in this study using the method of Backpropagation Neural Network and Genetic Algorithm to feature selection. On this panelitian proves that the method of Backpropagation Neural Network with genetic algorithm as a selection of the best RMSE value generating features of 0115. While the test results with the method of Backpropagation Neural Network without Genetic Algorithms as a value generating features selection RMSE 0152.