ANALYSIS OF LINEAR REGRESSION ALGORITHM TO PREDICT ADMINISTRATION, PRODUCTION AND DELIVERY TO ACCURACY OF DELIVERY OF PRODUCTS IN COSMETIC INDUSTRY, PT CEDEFINDO
Windu Gata, Hafifah Bella Novitasari, Ridan Nurfalah, Rani Hernawati , Muhammad Jafar Shidiq
STMIK Nusa Mandiri, Jakarta, Indonesia
Abstract
PT Cedefindo is engaged in the cosmetics contract manufacturer where the accuracy of product delivery is very important in order to maximize service to consumers, so information is needed to support the performance from order to delivery of the product. This study will provide predictions of product delivery based on the results of analysis of existing production history data. In addition, this prediction will prevent problems in administration, production and delivery. The data mining algorithm used is linear regression to predict administration, production, and delivery of product delivery accuracy. The predictive value produced is a constant of 6.105, plus 0.925 times the number of administrative days, plus 0.64 times the number of days of production, plus 0.191 times the number of days of delivery. The results of the test formula produced there is a difference of 10 days from the target and prediction.
Keywords: data mining, prediction, manufacturer, linear regression
Topic: Informatic and Information System