Prediksi Penjualan Tanaman Hias menggunakan Regresi Linier Berganda dengan Perbandingan Eliminasi Gauss dan Cramer

Sales prediction is a crucial element in the ornamental plant business to support inventory planning and marketing strategies. Our research aims to compare the Gauss and Cramer elimination methods (determinant matrix) in multiple linear regression to assess the accuracy of sales prediction. Gauss elimination is effective for systems of large size, while the Cramer method is more consistent in handling systems of linear equations that have correlated variables. The dataset used consists of 212 data points, including unit price as the dependent variable and stock, quantity sold, and total revenue as the independent variables. The accuracy was compared using Mean Absolute Percentage Error (MAPE) due to its ability to measure the error relative to the true value. Our findings show that the Cramer method has a MAPE of 21%, which is lower than Gauss elimination with a MAPE of 40%, making it more accurate in sales prediction. With a more precise method, business owners can optimize inventory management, set prices more efficiently, and devise data-driven marketing strategies. Our results also provide insights for other sectors that use predictive analytics to improve business decision-making.