Open AccessImplementation of Water Conditions in Soil with Artificial Neural Network Method using Backpropagation
Author(s) -
Toppan Sintio,
Steven Steven
Publication year - 2021
Publication title -
journal of computer networks, architecture and high performance computing
Language(s) - English
Resource type - Journals
ISSN - 2655-9102
DOI - 10.47709/cnahpc.v3i2.950
Subject(s) - backpropagation , artificial neural network , computer science , rprop , data mining , data collection , test data , artificial intelligence , machine learning , statistics , mathematics , time delay neural network , types of artificial neural networks , programming language
In agriculture and plantations, the land is an important thing, but sometimes the soil needs to be measured for its fertility, so measuring instruments are used. In this study, the authors tried to collect data for measuring using the Backpropagation method to determine the prediction of fertility in the soil. The backpropagation method is used to predict and also Backpropagation is a Neural Network algorithm. In using this method, input is a sensor that will take data in the form of soil moisture, pH when wet, and pH when dry, followed by this method which processes the data to be generated. The results of research with the Backpropagation algorithm get 80% accuracy of the 10 test data used for test results. The results tested initially were not as expected but with several trials, it was almost as expected but needed to be further developed. With the hope that there are people who can develop better for more knowledge and hopefully it can be useful for more. The suggestion in this is for readers who want to develop their suggestions to collect more data from this research to get more satisfying results. If the data is not more efficient, more efficient or accurate methods or means of data collection are expected to be used.