Machine-Learning Approach to Optimize SMOTE Ratio in Class Imbalance Dataset for Intrusion Detection
Table 9
Time complexity of the proposed methods.
Num. of the procedures of the proposed methods
Complexity (Big-O)
Worst case
1
// is the number of experiments O()
O()
2
// h is the number of 100 tuples which are randomly generated. O(h)
O(h)
3
O(h) + svm_train + O(h)
O(n2·h)
4
// The time complexity of svr_train is identical to svm_train. O(h) + svr_train
O(n2·h)
5
// k is the number of 1,000,000 tuples which are randomly generated. // If an algorithm does not depend on n, which is a symbol of amounts of data, then the algorithm has constant complexity or symbolized by O(1) [23]. Therefore, the time complexity of svr_test is identical to O(1). k∗ svr_test