23 Kas 2024 Cumartesi
Determination of optimization of different release parameters of Oenopia conglobata (Coleoptera: Coccinellidae) used in biological control of pistachio psylla Agonescena pistaciae (Homoptera: Psyllidae) with ecological parameters by different machine learning methods
Abstract : In this study, it was aimed to determine the regression model of the release parameters of the predator Oenopia conglobata' (Coleoptera: Coccinellidae) used in the biological control of the pistachio pest Agonescena pistaciae (Hompptera: Psyllidae) at different doses (25, 50 and 100 adult individuals) in 8 different locations depending on ecological parameters such as wind, temperature and humidity. For this purpose, 7 different ML regression algorithms were tested and KNNR and FNNR methods gave the best results in beneficial and pest prediction. These methods were the most appropriate methods for predicting the harmful changes depending on the beneficial releases in all three release numbers. Compared to these methods, MLR and MPR methods showed lower performance. The KNRR method was the most efficient method in terms of release optimization of beneficial insects and prediction in the field with its closeness to real values, and a prediction was made that this method can be used in other beneficial insect releases. This method was found to be promising in terms of providing ecologically based recommendations for the prediction of different ecological factors, especially in beneficial insect releases. Especially in the RFR method, pest, wind, habitat and release number differences were distributed on the same linear equation. The results of the study suggest that the classical biological control approach based on release to be used in pest control should be based on ecological parameters.