PENERAPAN METODE SUPPORT VECTOR MACHINE (SVM) DALAM KLASIFIKASI KUALITAS PENGELASAN SMAW (SHIELD METAL ARC WELDING)

Abstract

Quality control of a product must be maintained, so that consumers feel satisfied in using the products produced. One way that can be done by the industrial world is efficiency in product quality classification. A very good classification method compared to conventional methods, is the Support Vector Machine (SVM) method. The Support Vector Machine method is a supervised learning classification method. The SVM method is an algorithm that works using nonlinear mapping to change the original training data to a higher dimension. The purpose of the research is to obtain a classification model that has high accuracy or small errors in welding quality classification. The target of the researcher is to produce a control device to maintain effective and efficient welding quality. This research is a study that uses actual data, using the second data obtained from March 2018 to May 2018. The results of testing the model using a quadratic function kernel shows the accuracy of 96.2%, and testing using test data shows the results of accuracy 98% using a quadratic function kernel.