Investigations on the Influence of Variations in Hidden Neurons and Training Data Percentage on the Efficiency of Concrete Carbonation Depth Prediction with ANN

Uwanuakwa, Ikenna D.; Akpınar, Pınar

doi:10.1007/978-3-030-35249-3_128

Cited by 5 publications

(3 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Untuk prediksi karbonasi beton diperoleh nilai koefisien relasi (R) sebesar 0,99 dengan variasi hasil persentase 30-55% dengan arti, bahwa terdapat peningkatan nilai R secara signifikan dari yang diamati sebesar 60-80%. Selain hal itu, berdasarkan pengamatan ditunjukkan, bahwa variasi neuron tersembunyi antara 5-25 dihasilkan perubahan akurasi prediksi kurang signifikan, baik nilai R, MSE, maupun persentasenya [16].…”

Section: Pendahuluanunclassified

Algoritma Fungsi Perlatihan pada Machine Learning berbasis ANN untuk Peramalan Fenomena Bencana

Wanto

Defit

Windarto

2021

RESTI

View full text Add to dashboard Cite

Research has been carried out with several training functions using standard backpropagation methods, One-Step Secant (OSS), and Bayesian regulation. The purpose of this study was to (i) analyze the Performance accuracy (Performance) of the standard backpropagation method and (ii) optimize the training function with the One-Step Secant (OSS) and Bayesian regulation methods to obtain comparison results of the three methods in the search for the best results implementation of disaster phenomenon forecasting data. The research method is based on quantitative methods with times-series data on disaster phenomena in Indonesia over the last ten years (2011-2020) which were analyzed using two network architecture models, namely 4-8-1 and 4-10-1. The results showed that the 4-8-1 architectural model with the Bayesian regulation training function method was able to optimize quite well through accelerating training time and resulted in a low MSE measurement, although not the lowest with an epoch value of 197 iterations and a Performance of 0.0148480766. The lowest epoch value is generated by the OSS method, but it Performs poorly. The best Performance is produced by the standard backpropagation method with the traingd training function, but the training process for achieving convergence is also too long. In general, it can be concluded that the 4-8-1 architectural model with Bayesian regulation can be used to predict (predict) the phenomenon of natural disasters in Indonesia because the training time to achieve convergence is not too long and Performs exceptionally well.

show abstract

Section: Pendahuluanunclassified

Algoritma Fungsi Perlatihan pada Machine Learning berbasis ANN untuk Peramalan Fenomena Bencana

Wanto

Defit

Windarto

2021

RESTI

View full text Add to dashboard Cite

show abstract

“…Furthermore, observations show that changing the number of hidden neurons from 5 to 25 resulted in a less significant change in prediction accuracy, both in terms of R, MSE, and percentage values. [16].…”

Section: Introductionmentioning

confidence: 99%

Artificial Neural Network Model for Forecasting Natural Disasters: Polak-Ribiere and Powell-Beale Comparison

Rianti

Yenila

Ariana

et al. 2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

The prediction problem is an interesting topic to be discussed today. The many predictive methods used to solve problems have become an obstacle for researchers and academics alike. This study aimed to analyze the ability of the ANN prediction method using the Polak-Ribiere and Powell-Beale conjugate gradients. The dataset used for the analysis is disaster times-series data in Indonesia for the last ten years (2011-2020). Data obtained from the Indonesian Disaster Geoportal sourced from the National Disaster Management Agency can be seen on the infographic menu on the website https://gis.bnpb.go.id/. The results obtained based on the analysis that has been carried out, that the 4-10-1 architectural model with the Powell-Beale Conjugate gradient method can produce lower MSE Testing/Performance than the Polak-Ribiere Conjugate gradient method, another advantage is faster time. And fewer iterations. So it can be concluded that based on comparing these two methods, the Conjugate gradient Powell-Beale method with the architectural model 4-10-1 can be used for forecasting/predicting natural disasters because it is a better method.

show abstract

“…Dimana nilai (R) 0,99 dan persentasi hasil variasi 30-50% yang berarti nilai R meningkat signifikan dari yang diamati 60-80%. [38] Penelitian selanjuntnya dilakukan untuk meramalkan radiasi gelombang panjang teluk benggala menggunakan lima fungsi pelatihan yang berbeda, yaitu: kuasi-Newton Broyden Fletcher Goldfarb Shanno (BFGS), garis potong satu langkah, gradien konjugasi dengan Beale-Powell restart, Levenberg-Marquardt, dan penurunan gradien. dalam penelitian menggunakan analisa performance (Root mean square error, korelasi koefisien, akurasi prediksi) dan analisa parameter untuk mengetahui performansi fungsi pelatihan.…”

unclassified

Analisis Optimasi Fungsi Pelatihan Machine Learning Neural Network dalam Peramalan Kemiskinan

Sitanggang¹,

Defit

Ramadhan³

2021

JEPIN

View full text Add to dashboard Cite

Banyak metode fungsi pelatihan dalam Machine Learning Neural Network yang digunakan dalam menyelesaikan masalah komputasi yang berkaitan dengan prediksi. Fungsi pelatihan yang digunakan pada Machine Learning metoda algoritma backpropagation dapat menghasilkan prediksi yang berbeda, yang dipengaruhi oleh parameter dan data yang digunakan. Tujuan dari penelitian dilakukan untuk menganalisa performance dan keakuratan algoritma backpropagation standard serta mengoptimalkan fungsi pelatihan dengan algoritma Bayesian Regulation, dan One Step Secant. Dalam proses analisis, penelitian ini menggunakan Dataset jumlah kemiskinan di Indonesia dalam jangka waktu 12 tahun (tahun 2009 - 2020) yang terdiri dari 34 provinsi. Data diperoleh dari website Badan Pusat Statistik (BPS) Indonesia https://www.bps.go.id/. Berdasarkan pelatihan, pengujian, dan analisa yang dilakukan diperoleh hasil dari penelitian, bahwa model jaringan 5-9-1 menggunakan fungsi pelatihan Bayesian Regulation mampu melakukan optimasi yang lebih baik dengan percepatan waktu pelatihan, MSE Pengujian, Performance lebih rendah dibandingkan denga 2 metode yang lain, dengan demikian disimpulkan bahwa model jaringan 5-9-1 menggunakan algoritma Bayesian Regulation dapat digunakan untuk prediksi kemiskinan di Indonesia.

show abstract

Investigations on the Influence of Variations in Hidden Neurons and Training Data Percentage on the Efficiency of Concrete Carbonation Depth Prediction with ANN

Cited by 5 publications

References 22 publications

Algoritma Fungsi Perlatihan pada Machine Learning berbasis ANN untuk Peramalan Fenomena Bencana

Algoritma Fungsi Perlatihan pada Machine Learning berbasis ANN untuk Peramalan Fenomena Bencana

Artificial Neural Network Model for Forecasting Natural Disasters: Polak-Ribiere and Powell-Beale Comparison

Analisis Optimasi Fungsi Pelatihan Machine Learning Neural Network dalam Peramalan Kemiskinan

Contact Info

Product

Resources

About