Carnivorous Plant Algorithm and BP to Predict Optimum Bonding Strength of Heat-Treated Woods

In this study, the CPA algorithm was used to optimize a BP neural network model predict bond strength and surface roughness of heat-treated wood. The trained optimized using MATLAB software. results network, random forest algorithm, CPA-BP were compared. show that CPA-optimized has better R2 compared conventional model. After model, value increased by 8.1%, MAPE decreased 3.74%, MAE 33.91% in prediction strength. values 3.02% 20.47%, respectively, predicting mean maximum roughness. indicate is reliable wood Using can also reduce required experimental cost.