Akurasi Keuntungan Penjualan Bata Menggunakan Metode Mamdani (Studi Kasus: PD. Bahan Bangunan Cahaya Terang 9 Malang)

Abstract

This research was conducted based on the problems that occur in PD. Bahan Bangunan Cahaya Terang 9 Malang. The buildup of brick stock due to a mismatch with the number of consumer orders causes a decline in brick quality and sales profits. Therefore, the Fuzzy Mamdani Inference System method was applied to brick sales data from January 2023 to December 2023 to solve existing problems. Data processing was carried out manually and through Matlab programming. In this research, two data were processed using the Mamdani method, namely normalized data and denormalized data. Normalized data is used for comparative analysis between program results and calculation results. Meanwhile, denormalized data is used to test the level of accuracy of the research prediction model. There are several stages in applying the Fuzzy Mamdani method to brick sales data, namely fuzzification, application of Zadeh operations, application of implication functions, rule composition, and defuzzification. After that, an accuracy level test will be carried out using the Mean Absolute Percentage Error (MAPE) method and calculating profits in one period. Based on the results of data processing, in December 2023 the results showed that there would be 5,944 brick units that would be sold with a total profit of IDR. 1,210,000.00, while the test results for the accuracy level of the Fuzzy Mamdani method were 3.27%. Based on the number of brick sales based on the program results, it is not much different or almost accurate to the actual sales number. It can be concluded that the Fuzzy Mamdani method can be applied to analyze the accuracy of profits from brick sales in PD. Bahan Bangunan Cahaya Terang 9 Malang.