Detecting Denial of Service attacks using machine learning algorithms

Journal of Big Data

Table 6 Summary of the results for the experimental model

	Total instances	Correctly classified instances	Accuracy	Precision	Recall	Mean Absolute Error	Machine Learning Algorithm used	Class
Training data	14,063	14,063	100	1.000	1.000	0	Logistic Regression	Attack
	14,063	13,929	99.04	1.000	0.981	0.007	Naïve Bayes	Attack
	14,063	14,063	100	1.000	1.000	0	Logistic Regression	Normal
	14,063	13,929	99.04	0.981	1.000	0.007	Naïve Bayes	Normal
Test data	5425	5417	99.85	1.000	0.997	0.0015	Logistic Regression	Attack
	5425	5385	99.26	1.000	0.985	0.0061	Naïve Bayes	Attack
	5425	5417	99.85	0.997	1.000	0.0015	Logistic Regression	Normal
	5425	5385	99.26	0.986	1.000	0.0061	Naïve Bayes	Normal
Validation Data	602	601	99.83	1.000	0.997	0.0017	Logistic Regression	Attack
	602	594	98.67	1.000	0.974	0.0163	Naïve Bayes	Attack
	602	601	99.83	0.997	1.000	0.0017	Logistic Regression	Normal
	602	594	98.67	0.974	1.000	0.0163	Naïve Bayes	Normal