Drug Sales Prediction with ACF and PACF Supported ARIMA Method

摘要

With the technology developing day by day, data collection and storage becomes easier. It is possible to find univariate time series data based on a certain time period in fields such as finance, energy, meteorology and health. Estimating the future values of these series are important in many aspects such as planning and taking precautions. In this study, weekly warehouse exits of a pharmaceutical company were made into a time series on a drug basis and time series estimations were made with ARIMA model supported by ACF, PACF. The results obtained were compared with Facebook's Prophet library and it was observed that the developed model gave better results.