Implementation of Predicting the Availability of Chicken Eggs on Christmas Day Using Artificial Neural Network Backpropagation

Abstract

Prediction can be called a science that is used to predict events that are likely to occur in the future based on past events. One of the other prediction methods in circulation is Backpropagation Neural Network. Backpropagation Neural Network (BPNN) is a Neural Network (NN) that is forward in nature and does not have a loop through which signals flow from input neurons to output neurons. This research aims to determine a prediction of egg supply in 2023, especially during Christmas in Manokwari district to meet market and customer needs. By analyzing the availability of egg supplies in the city of Manokwari from January 2018 to December 2022. From the methods used in this research, starting from data collection methods as well as variables and research stages which include the data collection process, data sharing, then training and data testing and validation crosswise, the prediction pattern for the number of egg stocks is 12-16-1, where there are 12 variables in the input layer, then 16 variables in the hidden layer, 1 variable in the output layer, the learning rate value is 0.9 and the value the momentum is 0.1, resulting in a prediction of egg stock in 2023, especially in December, of 131053 eggs. With a MAPE value of 27.4767%. with the results of a feasible prediction model value. With the predicted results, the number of egg stocks in 2023, especially in December (during Christmas celebrations) in Manokwari Regency is 131,053 eggs during December 2023.