Compiler Optimization Prediction with New Self-Improved Optimization Model

Users may now choose from a vast range of compiler optimizations. These optimizations interact in variety sophisticated ways with one another and the source code. The order which optimization steps are applied can have considerable influence on performance obtained. As result, we created revolutionary prediction model. Our model comprises three operational phases: training, feature extraction, as well exploitation. training step includes initialization formation candidate sample sets. inputs were then sent to extraction phase, retrieved static, dynamic, improved entropy features. extracted features optimized by exploitation employs an hunger games search algorithm best In this work, used Convolutional Neural Network predict based these selected characteristics, findings show that our innovative surpasses previous approaches.