Scholarlink

The sale of dairy products has always been in the top five with the highest sales value from 2021 to early 2024, although the monthly sales value of the dairy product is quite fluctuating. There was a problem in predicting demand that was less accurate compared to actual demand in the market, which then had a significant impact on the company's operations. The analysis of the forecast of the sale of dairy products has become crucial with the aim of avoiding a shortage or surplus of supplies that could affect the performance of the warehouse and the quality of the service. The objective of the study is to model the sales experience in the first five months of 2024 using the Support Vector Regression (SVR) method, analyze the accuracy of predictions, and create a dashboard of prediction results using Power Business Intelligence (Power BI) software, and verify and validate the research that has been done. Based on calculations with SVR, the best kernel obtained is the Linear kernel with a Mean Absolute Percentage Error (MAPE) value on the training data of 1.749% which belongs to the category of highly accurate predictions and the value of the determination coefficient of 0.98, whereas for the data testing, the result of the Mean Absolute Percentage Error (MAPE) of 0.843%. which falls into the very accurate forecasting category and the determining coefficient value of 0.794 so it can be concluded that the predictive model capabilities used can be effectively applied to the prediction of the sale of dairy products.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

The Support Vector Regression Method with the Grid Search Algorithm in Forecasting Sales of Milk Product

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Nailah Khalishah Auliyaanisa,
  • Rina Fitriana,
  • Elfira Febriani Harahap

摘要

The sale of dairy products has always been in the top five with the highest sales value from 2021 to early 2024, although the monthly sales value of the dairy product is quite fluctuating. There was a problem in predicting demand that was less accurate compared to actual demand in the market, which then had a significant impact on the company's operations. The analysis of the forecast of the sale of dairy products has become crucial with the aim of avoiding a shortage or surplus of supplies that could affect the performance of the warehouse and the quality of the service. The objective of the study is to model the sales experience in the first five months of 2024 using the Support Vector Regression (SVR) method, analyze the accuracy of predictions, and create a dashboard of prediction results using Power Business Intelligence (Power BI) software, and verify and validate the research that has been done. Based on calculations with SVR, the best kernel obtained is the Linear kernel with a Mean Absolute Percentage Error (MAPE) value on the training data of 1.749% which belongs to the category of highly accurate predictions and the value of the determination coefficient of 0.98, whereas for the data testing, the result of the Mean Absolute Percentage Error (MAPE) of 0.843%. which falls into the very accurate forecasting category and the determining coefficient value of 0.794 so it can be concluded that the predictive model capabilities used can be effectively applied to the prediction of the sale of dairy products.