EVFDT: An Enhanced Very Fast Decision Tree Algorithm for Detecting Distributed Denial of Service Attack in Cloud-Assisted Wireless Body Area Network

Table 5

Comparison of existing machine learning techniques with VFDT.

Features	VFDT-	CVFDT	OVFDT	EVFDT
Detection accuracy	Very low	Very low	Good; does not handle outliers	Excellent
Resource usage (time/memory)	Less time; more memory	More time in building two trees; requires additional memory	Less time; less memory	Having same time as VFDT but consuming very less memory space
Noisy data handling	Does not handle noisy data	Not appropriate under noisy data	HB fluctuation intensifies under noisy data; accuracy decreases	Handles noisy data efficiently
Tree size/pruning	Small tree size; no pruning	Same tree size as VFDT; no pruning	Small tree size; incremental pruning	Small tree size; iterative pruning
Computational resources	Consuming less resources	Consuming more resources by maintaining two trees	Consuming less resources	Consuming very less resources by cutting of HB outliers