Flight Ticket Prediction using Random Forest Regressor Compared with Decision Tree Regressor
NSN Rao, S. John Justin Thangaraj
Abstract
The objective of the research study is to predict and analyze the flight ticket fare using machine learning algorithm Random Forest Regressor in comparison with Decision tree Regression algorithm. The Random Forest regressor (N=10) and Decision tree regressor algorithm (N=10) with the estimated sample sizes are taken and the prediction accuracy was calculated by using two groups and a total of 20 samples taken for both algorithms in this work. The samples were processed as 10 per group using a G-Power value of 0.8, threshold 0.05% and CI at 95%. The prediction Accuracy of both the algorithms are identified as the Random Forest Regressor with 86.70% and the Decision Tree Regressor with 79.69%. The statistical significance difference between GradientBoosting Regressor and AdaBoostRegressor was found to be 0.00 in the 2-tailed test (p<0.05). After all the Procedures the Prediction of Flight Prices using Random Forest Regressor seems to appear more accurate while compared with the Decision tree Regressor.