Table 2 Symbols descriptions.

\(\mathcal{S}\)

Data stream

\(\dot{\mathbf{S}}\):

List: samples seen for each class

К:

Ensemble size

\(\overline{{{\mathbf{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle\cdot}$}}{R} }}}}\):

List: current imbalance ratio of each class

η

Feature subset size

\(\ddot{\Upsilon }\):

Learner in background ensemble

\(\mathcal{W}\)

Window size

\(\mathcal{F}\)

Feature

Ĭ:

Instance

ε:

Ensemble

Ψ:

Feature subset

\(\mathcal{B}\):

Background ensemble

\(\mathcal{F}{\mathbf{w}}\):

Warning level for change in feature data

\(\:{\mathbf{DIR}}_{\varvec{C}}\)

Dynamic imbalance ratio of class C

δ_w

Performance threshold for warning

\({{\bar{\mathbf{w}}}}\):

Class weight

\(\:{\mathbf{DIR}}_{\varvec{C}}\)

Cumulative imbalance ration of class C

\(\:\varvec{y}\):

Class label

\({\mathcal{F}}{\mathbf{d}}\):

Drift level for change in feature data

\(\mathcal{Z}\):

Number of classes

δ_d

Performance threshold for drift

Ρ:

Feature subset pool

\({\calligra{m}}\):

Number of features

\(\Upsilon ^{\prime}\):

Background Learner in main ensemble

\(\Upsilon\):

Learner in main ensemble