What question did this study set out to answer?

This research aims to develop a prediction model for carbon monoxide detection using artificial neural networks.

December 23, 2025JASIEK (Jurnal Aplikasi Sains Informasi Elektronika dan Komputer)0 citationsOpen Access

Prediksi Gas Karbon Monoksida dengan Jaringan Syaraf Tiruan berbasis Internet of Things

AHAlauddin Maulana HirzanUniversitas Semarang CMCharis MaulanaUniversitas Semarang SHSri HandayaniState University of Semarang

Key Points

This research aims to develop a prediction model for carbon monoxide detection using artificial neural networks.
Developed a prediction model using an Artificial Neural Network algorithm.
Evaluated four ANN models: ANN-5K, ANN-6K, ANN-4K, ANN-3K.
Compared the performance of ANN models against linear regression.
The ANN-5K model achieved the highest accuracy of 80.18%.
Other models had accuracies of 77.13% (ANN-6K), 66.44% (ANN-4K), and 53.14% (ANN-3K).
Linear regression had a lower accuracy of 57.50%, indicating the superiority of the ANN models.

Abstract

Carbon monoxide is a dangerous gas that can cause fatal effects in humans if inhaled in large quantities. To detect it, a model has been developed. This study proposes a prediction model using an Artificial Neural Network (ANN) algorithm to predict carbon monoxide. Of the four ANN models evaluated, the ANN-5K model showed the best performance with an accuracy of 80.18%, followed by ANN-6K with an accuracy of 77.13%, ANN-4K with 66.44%, and ANN-3K with 53.14%. When compared to linear regression, which only had an accuracy of 57.50%, the ANN-5K model was still superior. Thus, the proposed ANN-5K model proved to be more accurate and had a lower error rate compared to other models. The main contribution of this research is a prototype equipped with an ANN model to predict carbon monoxide gas

问 AI

Bookmark

View Full Paper