Research on the Intrusion Detection Model of Underwater Sensor Networks

Algorithm 3

SMOTE algorithm sampling steps.

Input: training sample S, k-nearest neighbor parameter k, number of minority samples n

Output: SMOTE sampling result

1: Calculate the k-nearest neighbor similar set of minority class sample ;

2: Randomly select the sample in the set;

3: Synthesize a new sample between and according to Equation (9);

4: Add a new sample to the training sample to get a new training sample S_new;

5: Use a classifier to classify S_new.