Download now Free registration required
Network Intrusions are critical issues in computer and network systems. Several intrusion detection approaches be present to resolve these severe problems but the major problem is performance. To increase performance, it is significant to increase the detection rates and reduce false alarm rates in the area of intrusion detection. The recent approaches use Principal Component Analysis (PCA) to project features space to principal feature space and select features corresponding to the highest eigenvalues, but the features corresponding to the highest eigenvalues may not have the optimal sensitivity for the classifier due to ignoring many sensitive features.
- Format: PDF
- Size: 576.7 KB