New Approach Motion Determination in Pencak Silat Sport using Self Organizing Map

Pencak Silat is a sport that not only pays attention to the strength factor but also cultural and spiritual elements. This is what makes Pencak Silat one of the cultural heritages of Indonesia. In order not to become extinct, efforts are needed, one of which is given a touch of technology in the practice of Pencak Silat matches. The aim of this research is implementing the Self Organizing Map in determining the type of Pencak Silat movement. The movements that will be identified are punches, kicks, and defence. This is important because the determination of the score depends on the movement of punches, kicks, and defence did by the athlete. This research consists of several stages, namely collecting data on smart pirates, processing data statistically, carrying out the learning process using a Self Organizing Map, conducting the testing stage, and conducting the evaluation stage. Data collection was carried out primarily by involving 6 people (there were 3 pairs of matches) to perform limited punches, kicks, and parries. Each person does 10 punches, 10 kicks, and 10 parries. So the total data is 180 data. The method used is Self Organizing Map with a learning rate of 0.8 and the number of epochs in the learning process is 1000 epochs. The results show that the system has determined the motion well, supported by the accuracy value during the testing process. The accuracy value obtained is 74.444% for the first and second experiments, and 99.444% for the third experiment.