Cascade-forward neural network (CFNN) for biomass heating value prediction
Cascade-forward is a class of artificial neural network that has the same characteristics with feed-forward neural networks. However, cascade-forward neural network connects the input and each earlier layer with the following layers. In other words, in Cascade-Forward Neural Network (CFNN), every ne...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | en |
| Published: |
2023
|
| Online Access: | http://eprints.utem.edu.my/id/eprint/27910/1/Cascade-forward%20neural%20network%20%28CFNN%29%20for%20biomass%20heating%20value%20prediction.pdf http://eprints.utem.edu.my/id/eprint/27910/ https://pubs.aip.org/aip/acp/article-abstract/2727/1/030010/2895047/Cascade-Forward-Neural-Network-CFNN-for-biomass?redirectedFrom=PDF |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Cascade-forward is a class of artificial neural network that has the same characteristics with feed-forward neural networks. However, cascade-forward neural network connects the input and each earlier layer with the following layers. In other words, in Cascade-Forward Neural Network (CFNN), every neuron in the input layer is connected to every neuron in the hidden and output layer. Cascade-Forward Neural Network is practically useful for any type of input-to-output mapping. The benefit of Cascade-Forward Neural Network is that it can provide the nonlinear input-output relationship without removing its linear relationship. In this study, Cascade-Forward Neural Network was utilised to predict heating value of biomass. The proximate analysis of 350 samples of biomass was used. To examine the prediction accuracy of the model, six parameters were examined. Results showed that, Cascade-Forward Neural Network trained with Levenberg-Marquardt backpropagation algorithm has successfully predicted biomass heating values with R, R2, MAD, MSE, RMSE and MAPE being 0.9640, 0.9293, 0.7786, 1.0642, 1.0316 and 4.4776, respectively. |
|---|