Abstract:
Sales are activities that need attention within the enterprise, since the end of a
production is the sale of goods. The availability of stock of goods depends on the results
of sales, the higher the sales figures, the greater the profit obtained by the company. In
order to maintain sales stability, several companies use data mining to help solve
production problems. This study aims to process transaction data in the Computer
Square store in order to predict future sales using the C4.5 algorithm method. The C4.5
algorithm is an algorithm that has the function of predicting an event in the future using
previous data. Predictions are very profitable for companies to maintain stock stability
to meet sales products that sell well and are not selling well in the market. Using
decision trees, the C4.5 algorithm is able to predict research results with high accuracy.
This study presents 23 rules with the conclusion that 11 rules state that they are not in
demand and 12 rules state that they are selling well. Data analysis using this algorithm
has an accuracy rate of up to 79.16%.