Proceedings of the 1st International Conference on Science and Technology for an Internet of Things, 20 October 2018, Yogyakarta, Indonesia

Research Article

Implementation of Backpropagation Neural Network Method in Classification System of Timeliness of Graduation

Download38 downloads
  • @INPROCEEDINGS{10.4108/eai.19-10-2018.2282527,
        author={Yanuar Nurdiansyah and Diksi Media and Fadhel Akhmad Hizam},
        title={Implementation of Backpropagation Neural Network Method in Classification System of Timeliness  of Graduation},
        proceedings={Proceedings of the 1st International Conference on Science and Technology for an Internet of Things,  20 October 2018, Yogyakarta, Indonesia},
        publisher={EAI},
        proceedings_a={ICSTI},
        year={2019},
        month={4},
        keywords={data mining classification artificial neural network backpropagation neural network method},
        doi={10.4108/eai.19-10-2018.2282527}
    }
    
  • Yanuar Nurdiansyah
    Diksi Media
    Fadhel Akhmad Hizam
    Year: 2019
    Implementation of Backpropagation Neural Network Method in Classification System of Timeliness of Graduation
    ICSTI
    EAI
    DOI: 10.4108/eai.19-10-2018.2282527
Yanuar Nurdiansyah1,*, Diksi Media1, Fadhel Akhmad Hizam1
  • 1: Universitas Jember
*Contact email: yanuar_pssi@unej.ac.id

Abstract

Information Systems is one of the study programs at the University of Jember which was established since 2009. To earn a degree of bachelor, students must pass 144 credits of 4-5 years study period. Method for classifying using an artificial neural network of Backpropagation. Attribute used for classification is 9 attributes, namely Grade Point Average (GPA) from 1 to 6 semester, the number of credits taken, the last semester taking Student Study Service (KKN) and Internship (PKL). Class used for the classification is the timeliness of graduation. The implementation of this classification method is done by learning rate 0.1, 0.3, 0.5, 0.7, and 0.9 with the iteration limit of 1,000, 2,000, and 3,000. The highest accuracy is 98.82% for the 2000 and 3000, each with learning rate = 0.7 and 0.9 for the 2000th iteration and learning rate = 0.5, 0.7 and 0, 9 for the 3000th iteration.