Implementation of Neural Network Method for Air Quality Forecasting in Jakarta Region

Abstract

Abstract The quality of air can be influenced by the amount of pollution that occurs in an area. The city of Jakarta is ranked in the top ten as the nation’s capital with the worst air quality in the world. Poor air quality both inside and outside the room can have an impact on the emergence of various diseases and even death. For this reason, forecasting of air quality in the city of Jakarta, Indonesia is needed to anticipate the likely impact that will arise. In this study forecasting air quality using the neural network method in which this method has the advantage of being able to solve problems, especially large data samples and has been able to prove in handling non-linear problems. The data collection used is secondary data from the Environmental Service Office of DKI Jakarta Province as many as 2989 records with variables as determinants consisting of 5 of which PM10, SO2, CO, O3, NO2 and 1 output variable are good, moderate, unhealthy and very unhealthy. From the calculations result in this study it is known that the Neural Network method obtained an accuracy performance of 88.86% in which the Lubang Buaya area noted as the most unhealthy air quality.