Jurnal Riset Teknologi dan Inovasi Pedidikan (JARTIKA)
Vol. 2 No. 2 (2019): Juni

Analisis Tingkat Akurasi Model Backpropagation Dalam Prediksi Indeks Pembangunan Manusia (IPM) di NTB

Mila Haryati (Universitas Islam Negeri (UIN) Mataram)
Rosdiana Rosdiana (Universitas Islam Negeri (UIN) Mataram)
Khaerunnisa Khaerunnisa (Universitas Islam Negeri (UIN) Mataram)
Okta Widya Arsani (Universitas Islam Negeri (UIN) Mataram)
Mufti Wahyudi (Universitas Islam Negeri (UIN) Mataram)



Article Info

Publish Date
28 Sep 2021

Abstract

This study aims to determine the type of training in the most accurate Artificial Neural Network (ANN) in the Backpropagation method which was applied using matlab. In the simulation, the authors used the Human Development Index (HDI) data in the Province of West Nusa Tenggara (NTB) in 2010-2018 to predict the Human Development Index (HDI) data in West Nusa Tenggara (NTB) in 2019. Artificial neuralism with the Backpropagation method valued are Traingd, Traingdx, And Trainrp with the accuracy of each type of training algorithm using MSE. Based on the simulation results, the data uses Traingd, Traingdx, And Trainrp which can be used to assess the most popular regression in graph 1 Traingd with a value of R = 0.99744. Producing output from Traingd results Caused by the Human Development Index (HDI) data in 2019. The results of the predictions of ten districts / cities in 2019 in West Nusa Tenggara are 66,791, 65.2014, 65.2769, 67.0017, 67.3496, 66.0761, 72.2184, 64.8257, 78.3422, 76,618 and with error values of 0.29092, -0.13139, -0.07693, -0.73169, -0.53963, -0.50668, -1.8084, -0.68567, 0.38785, -1.768.

Copyrights © 2019






Journal Info

Abbrev

jartika

Publisher

Subject

Education Languange, Linguistic, Communication & Media Mathematics Social Sciences Other

Description

Focus and Scope: (1) Theoretical studies, research, and development of teaching materials, activities, models, or learning media in the field of education including mathematics and natural sciences (physics, chemistry, biology, science, ICT, and its applications); (2) social science, history, ...