Prediksi Ketebalan Powder Coating Menggunakan Algoritma SVM Dan Naïve Bayes

Abstract

Data Mining is a method that has been widely used to make scientific discoveries from a collection of datasets which so far have only been stored without further management. In the industrial world the use of data mining methods has helped with problems that are often found in the industrial field. Data mining helps in making predictions regarding thickness quality problems in a panel box product. Data mining is very useful for finding patterns in complex manufacturing data processing processes. Especially when we talk about consumers or service users of our product panels who want the panel to have good powder coating quality. This made the researchers conduct research to find the accuracy value which would later be used as a definite reference regarding the thickness of the powder coating. The results of this test the svm algorithm is better than naïve Bayes because the data in general can be categorized as a good result which has an accuracy of 97.60%, precision 99.56% and 96.03% recall. This res ult is an illustration for consumers to ensure that the panels to be purchased are of the best quality. By showing the data that has been processed, the consumer is sure that the purchase is really valid