Cuckoo search-ExtraTrees model for Radio-frequency power amplifier under different temperatures

Abstract This paper combines cuckoo search (CS) with ExtraTrees for the first time searching for the optimal hyperparameters and proposes the CS-ExtraTrees model. Based on the parasitic feeding behavior of cuckoo birds and the principle of Lévy flight enhancing random flying ability, CS can effectively search for the optimal parameters on a global scale. To verify the effectiveness of the CS-ExtraTrees model proposed in this paper, the CS-ExtraTrees model was used to model the measured data of the 2.5 GHz gallium nitride (GaN) power amplifier (PA). Compared with gradient boosting, random forest, and ExtraTrees, the CS-ExtraTrees proposed in this paper can significantly improve the modeling accuracy and modeling speed.