Discriminate primary gammas (signal) from the images of hadronic showers by cosmic rays in the upper atmosphere (background) with machine learning

Abstract

The grid search approach was used in this article to tune the hyperparameters of the Light Gradient Boosting Machine, Random Forest, Gradient Boosting, Extra Tree, Ada Boost, and Linear Discriminant Analysis algorithms for Gamma and Hadron classification. The findings of ROC and Precision-Recall curves were also discussed to assess the performance of algorithms in the Gamma and Hadron classification. with the light gradient boosting machine, it took 33 seconds to get an AUC value of 0.94 for the Gamma and Hardon classification. Also, the results of the Random Forest, the light gradient boosting machine, and the Linear Discriminant Analysis were all about the same.

Machine learning techniques for classifying dangerous asteroids

“…Eighty percent of the data is used in the 10-fold cross-validation and the whole of the remaining 20% of your data is unseen to the model (as the test set and validation set) [ [20] , [21] , [22] , [23] , [24] , [25] , [26] , [27] , [28] , [29] , [30] , [31] ] to test the model and another portion to test it. The prediction error from Equation One is then estimated using cross-validation as follows: …”