<s>
In	O
computer	B-General_Concept
science	I-General_Concept
,	O
incremental	B-Algorithm
learning	I-Algorithm
is	O
a	O
method	O
of	O
machine	O
learning	O
in	O
which	O
input	O
data	O
is	O
continuously	O
used	O
to	O
extend	O
the	O
existing	O
model	O
's	O
knowledge	O
i.e.	O
</s>
<s>
It	O
represents	O
a	O
dynamic	O
technique	O
of	O
supervised	B-General_Concept
learning	I-General_Concept
and	O
unsupervised	B-General_Concept
learning	I-General_Concept
that	O
can	O
be	O
applied	O
when	O
training	O
data	O
becomes	O
available	O
gradually	O
over	O
time	O
or	O
its	O
size	O
is	O
out	O
of	O
system	O
memory	O
limits	O
.	O
</s>
<s>
Algorithms	O
that	O
can	O
facilitate	O
incremental	B-Algorithm
learning	I-Algorithm
are	O
known	O
as	O
incremental	O
machine	O
learning	O
algorithms	O
.	O
</s>
<s>
Many	O
traditional	O
machine	O
learning	O
algorithms	O
inherently	O
support	O
incremental	B-Algorithm
learning	I-Algorithm
.	O
</s>
<s>
Other	O
algorithms	O
can	O
be	O
adapted	O
to	O
facilitate	O
incremental	B-Algorithm
learning	I-Algorithm
.	O
</s>
<s>
(	O
RBF	B-Algorithm
networks	I-Algorithm
,	O
</s>
<s>
the	O
incremental	O
SVM	B-Algorithm
.	O
</s>
<s>
The	O
aim	O
of	O
incremental	B-Algorithm
learning	I-Algorithm
is	O
for	O
the	O
learning	O
model	O
to	O
adapt	O
to	O
new	O
data	O
without	O
forgetting	O
its	O
existing	O
knowledge	O
.	O
</s>
<s>
Incremental	O
algorithms	O
are	O
frequently	O
applied	O
to	O
data	B-General_Concept
streams	I-General_Concept
or	O
big	B-Application
data	I-Application
,	O
addressing	O
issues	O
in	O
data	O
availability	O
and	O
resource	O
scarcity	O
respectively	O
.	O
</s>
<s>
Stock	O
trend	O
prediction	O
and	O
user	O
profiling	O
are	O
some	O
examples	O
of	O
data	B-General_Concept
streams	I-General_Concept
where	O
new	O
data	O
becomes	O
continuously	O
available	O
.	O
</s>
<s>
Applying	O
incremental	B-Algorithm
learning	I-Algorithm
to	O
big	B-Application
data	I-Application
aims	O
to	O
produce	O
faster	O
classification	B-General_Concept
or	O
forecasting	O
times	O
.	O
</s>
