Fig 1

Average proportion of failure vertices P as a function of ${\boldsymbol{\beta}} $ and ${\boldsymbol{\theta}} $ based on ML model and our method in BA and WS networks with $ \langle {\boldsymbol{k}}\rangle {\boldsymbol{= 8}} $ and N=1000"

Fig 2

Relationship between G and ${\boldsymbol{\beta }}$ for different $ {\boldsymbol{\theta }}$ under two methods in BA and WS networks with $ \langle {\boldsymbol{k}}\rangle {\boldsymbol{= 8}} $ and N=1000"

Fig 3

Relationship between ${{\boldsymbol{P}}_{{\boldsymbol{\max}} }}$ and ${\boldsymbol{\beta}} $ for different ${\boldsymbol{ \theta}} $ under two methods in actual networks"

Fig 4

P as a function of ${\boldsymbol{\beta}} $ and ${\boldsymbol{\theta}} $ based on two methods in typical networks when the load changes with $ \langle {\boldsymbol{k}}\rangle {\boldsymbol{= 8}} $ and N=1000"

Fig 5

Relationship between G and ${\boldsymbol{\beta}} $ for different ${\boldsymbol{ \theta}} $ under two methods in BA and WS networks when the load changes with $ \langle {\boldsymbol{k}}\rangle {\boldsymbol{= 8}} $ and N=1000"

Fig 6

Relationship between ${{\boldsymbol{P}}_{{\boldsymbol{\max}} }}$ and ${\boldsymbol{\beta}} $ for different $ {\boldsymbol{\theta}} $ under two methods in real networks when the load changes"

Fig 7

Relationship between ${\boldsymbol{Q}}$ and ${\boldsymbol{\beta}} $ for different ${\boldsymbol{ \theta}} $ under two methods in BA networks"

Fig 8

Relationship between ${\boldsymbol{Q}}$ and ${\boldsymbol{\beta}} $ for different $ {\boldsymbol{\theta}} $ under two methods in WS networks"