Interval Regression With Neuro-Fuzzy and Madaline Architecture for Prediction of Rice Production

DOI

10.2991/ictvt-17.2017.7How to use a DOI?

Keywords

interval regression; madaline architecture; neuro-fuzzy; rice production

Abstract

The paddy production interval value in Kubu Raya Regency can be predicted by using past values in conjunction with one of the neuro-fuzzy's methods, which is the interval regression model. Interval regression is a backpropagation-based method.There are two backpropagated-networks in this model, with one set to find a lower bound and the other set to find an uppe bound. The constructed system has the capability to receive input data in the form of statistics, which in this case is the rice production data of the last three periods, and then processing them using the interval regression model via the neuro-fuzzy. This Madelin Method uses three layers, with four input nodes, three hidden nodes, and one output nod, which then delivers the result in the form of a predictive value of future rice production values. Our primary goal is to predict next year's rice production. In this article, I used two machine prediction-based models, which is the interval regression model with neuro-fuzzy and the madaline.This research used 3 time periods to determine the best parameters for accuracy. To evaluate the performance of this system, we used the Mean Squared Error (MSE) measurement. The results gained indicated that the performance of the interval regression model with neuro-fuzzy worked better than the madaline with a learning parameter value of 0,09. The benefits of our method is that the regression interval method with neuro-fuzzy is one of the strongest machine prediction model in the world.

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Fatma Agus Setyaningsih
PY  - 2017/09
DA  - 2017/09
TI  - Interval Regression With Neuro-Fuzzy and Madaline Architecture for Prediction of Rice Production
BT  - Proceedings of the International Conference on Technology and Vocational Teachers (ICTVT 2017)
PB  - Atlantis Press
SP  - 35
EP  - 40
SN  - 2352-5398
UR  - https://doi.org/10.2991/ictvt-17.2017.7
DO  - 10.2991/ictvt-17.2017.7
ID  - Setyaningsih2017/09
ER  -