<![CDATA[Decision-making errors are fatal in company operations and often result in reduced profits. In making the right decisions, in-depth internal and external data analysis is required. One effort to increase decision-making accuracy is to apply a forecasting algorithm for existing transaction data. This research uses data mining and Forecasting for car repair companies. The information system being built is expected to be able to display a list of satisfied customer data, group transaction data based on the level of damage, and provide forecasts for the number of panels, number of units, and total revenue for the next period. From this data, it will be easier for management to determine the number of workers, additional tools, and stock of materials for the vehicle repair process. This information system was built by applying three algorithms, namely C4.5, K-Means, and Holt Winter Multivariate. The C4.5 algorithm will help group customer types, K-Means will help cluster vehicle damage levels, and Holt-Winters will be used for experience. The tests carried out included testing the accuracy of each algorithm, the black box system, and the system time so that the accuracy results were 100% for the C4.5 classification algorithm and 92.28% for the accuracy of K-Means which was tested with the purity measure. Meanwhile, the MAPE value for forecast testing was between 9 – 11% for each forecast.]]>
