Pengaruh Smoothing Data Terhadap Hasil Prediksi Volume dan Ritasi Sampah di Kota Bandung Menggunakan Metode Regresi Linear

The waste problem is very important in big cities, especially Bandung. The population, people's lifestyles and waste management that has not been carried out professionally are a challenge in itself. One of the preparatory steps to deal with the waste problem is to predict the development of waste volume. In this study, a statistical time series approach, namely the linear regression method, is used to predict the volume and transportation of waste in the city of Bandung. In the prediction process, data processing before being used in the prediction process plays an important role, one of which is data smoothing. A process to smooth the data using the moving average method with intervals of 2.3 and 4 and moving averages with weights of 242 and 12421 will be used to see its effect on the prediction results. Scenario 1 of the data used is all monthly data in the dataset time range and scenario 2 uses the same month's data for each year to predict the results of the month in the following year. The results of the volume and transportation predictions of waste between the best results from the Smoothing method and the results without going through the data smoothing process in scenario 1 show less significant results, namely less than 1%, while in scenario 2 it shows quite significant results, namely around 10% when compared to the actual data. Patterns and data ranges affect the final result from scenarios above.