Institusion
Institut Teknologi Perusahaan Listrik Negara
Author
SAINUDDIN, SHAFIRA AULIA RAMADANI
Aziza, Rosida Nur
Ningrum, Rahma Farah
Subject
Teknik Informatika
Datestamp
2023-05-31 07:50:38
Abstract :
Forecasting is an obligatory activity that is used to predicting future demand. In the
Bungsu pharmaceutical company there are 20-50 sales transactions that must be
recorded on a daily basis. As a results, to construct a drug data sales forecasting system
in a pharmaceutical, pharmacists have complexities estimating the amount to order of
each type in the following month. As a result, to construct a drug data sales forecasting
system in a pharmacy, the researcher conducted an ongoing system analysis process.
The Single Exponential Smoothing method is considered the most suitable for
forecasting drug sales data since it is a forecasting method on moving averages by
providing weights which are unchallenging to analyze. Forecasting analysis is done by
looking for the best alpha (?) value so as to produce an accurate forecast. The use of
data from the last 10 months as a reference data for recording the past for forecasting
experiments for the following month. Experiments with different alphas produced the
most accurate value, namely ?=0.1 with the smallest error rate, namely the smallest
MSE=358, 041 and MAPE=22%.