Random Forest and Stochastic Gradient Tree Boosting Based Approach for the Prediction of Airfoil Self-noise

Abstract Due to the complex nature of turbulent fluid flow, which is the source of airfoil self-noise generation, prediction of airfoil self-noise is a challenging problem, crucial for designing silent airfoils. In this paper, models for prediction of airfoil self-noise using two different algorithms namely Random Forest and Stochastic Gradient Tree Boosting have been proposed and tested on the NACA0012 airfoil data set, published by NASA. The parameters used to build these models were studied extensively and 10-fold cross validation was performed on the data set. The models predict with significantly higher accuracy compared to results reported by earlier computational models.