<![CDATA[This research was conducted to analyze the sales pattern of Indomie products at the UIN Sunan Kalijaga Yogyakarta Student Cooperative and predict future product prices. The data used is daily sales data from January to June 2023 with a total of 599 data into five clusters with the number of items cluster 0 consists of 32 items, cluster 1 consists of 409 items, cluster 2 consists of 102 items, cluster 3 consists of 48 items, and cluster 4 consists of 8 items. The methods used are K-Means for clustering and Support Vector Regression (SVR) for price prediction. The results of the K-Means analysis grouped the products into five clusters with different characteristics. In the Support Vector Regression (SVR) method, initially it has an accuracy rate of 70% with a fairly high Mean Squared Error (MSE) and Mean Absolute Error (MAE). After cleaning the data from outliers and tuning the hyperparameters, the model accuracy increased to 99%, showing a significant improvement in the model's predictive ability.]]>
Copyrights © 2024
