The Application of Artificial Neural Network Method in Forecasting the Number of Pregnant Woman Visits (K4)

Abstract

Forecasting is a systematic attempt to predict future events usingpast data, based on scientific and qualitative methods. For the maternal health program, forecasting is important as its process consists of planning, targetting and achievement. Based on data from the Ministry of Health, the quality of antenatal care in Indonesia was still low (87.48 percent) compared to that of the national target (95 percent). This study aims to apply the methods of artificial neural network in predicting the antenatalcare (K4). This applied research used a descriptive method with secondary data in the form of monthly antenatal care visits (K4) from the year of 2012 to2015 obtained from the Provincial Health Office of East Java, with a case study in Bondowoso. The forecasting result in 2016 based onthe 12-4-1 network architecture was 9533.5698, with the value of Mean Square Error (MSE) of 3091.84404. The average percentage of errord based on a comparison with the actual data is 0.1854 or reaching the accuracy of 99.81 percent. The conclusion of this study is that a neural network has a low error value and a high accuracy.Therefore, forecasting results can be used as an input in the planning program.