Discussion of a higher diploma thesis in the College of Computer Science and Mathematics - Department of Computer Science entitled (Pixel-based Machine Learning Algorithms for Faster Classification)
On Sunday, May 29, 2022, the College of Computer Science and Mathematics, University of Mosul, discussed the Higher Diploma Thesis (Pixel-based Machine Learning Algorithms for Faster Classification) by student Nour Rafea Abdel-Baqi, under the supervision of Assistant Professor Dr. Muhammad Gajan Younis
The thesis presented by the student dealt with testing the size of decision trees in the random forest algorithm for machine learning, where the random forest is a computationally efficient technique that can work quickly across a large data set.

The study dealt with the use of a data set (Oxford-IIIT Pet data set), which is a data set consisting of (349) images of cats from (13) different breeds. Random points were identified from the training data set to build decision trees and then choose the ideal number of trees that Leads to performance gains (in terms of accuracy of results and time spent).
The study aims to test a number of decision trees in the random forest algorithm, and it has been clarified through the graphs that increasing the number of trees in the random forest does not always mean that the performance of the forest will go towards the better. It was also concluded that the number of random forest trees 20 could be considered an ideal tree size in terms of accuracy of results and speed of implementation in relation to the database that was adopted in this study.

Also Read