Skripsi
PERBANDINGAN 1 DIMENSIONAL CONVOLUTIONAL NEURAL NETWORKS (1DCNN) DENGAN DECISION TREE DALAM MENDETEKSI PELANGGARAN LALU LINTAS KENDARAAN KOTA PALEMBANG
This research aims to compare two methods, namely 1 Dimension Convolutional Neural Networks (1DCNN) and Decision Tree, in detecting traffic violation rates in the city of Palembang. This study also uses You Only Look Once version 8 (YOLOv8) to count the number of vehicles to be detected based on video recordings, resulting in a model and obtaining an F-1 confidence score of 82%. The 1 Dimension Convolutional Neural Networks method achieved a good accuracy rate of 85%. Meanwhile, the Decision Tree method achieved an accuracy rate of 90%, indicating very good performance in detecting violation rates. Therefore, it can be concluded that the results of this study show that YOLO can detect objects with an accuracy of 82%, and Decision Tree performs very well in detecting violation rates with an accuracy of 90%, compared to the 1 Dimension Convolutional Neural Networks (1DCNN) method, which has an accuracy of 85%.