ANALYZING THE EFFECTIVENESS OF DECISION TREE AND NAÏVE BAYES CLASSIFIERS IN CLOUD COMPUTING ENVIRONMENTS

Abstract

CC has been growing rapidly in recent years due to its cost-effectiveness and efficient resource management. With this huge growth in the last ten years, many organizations have shifted their operations online for better reach, flexibility, and visibility. This research will do a comparison of two different machine learning techniques: Decision Tree and Naive Bayes Tree, based on precision and error detection. The output shows that NB Tree performed better with 97.05% correct predictions for an 80/20 split, 96.09% for a 70/30 split, and 96.78% for a 10-round cross-validation. But NB Tree is slower and took approximately 1.01 seconds because it involves more complexity in its algorithm. On the other side, the accuracy of the Decision Tree (J48) is comparatively low, at 96.78%, 95.95%, and 96.78% for the splits mentioned above, while taking significantly lesser time of 0.11 seconds. NB Tree outperformed J48 by about 0.9% in terms of precision, whereas J48 completed the analysis in about 9 seconds faster than NB Tree.