<s>
In	O
reinforcement	O
learning	O
(	O
RL	O
)	O
,	O
a	O
model-free	B-Algorithm
algorithm	O
(	O
as	O
opposed	O
to	O
a	O
model-based	O
one	O
)	O
is	O
an	O
algorithm	O
which	O
does	O
not	O
use	O
the	O
transition	O
probability	O
distribution	O
(	O
and	O
the	O
reward	O
function	O
)	O
associated	O
with	O
the	O
Markov	O
decision	O
process	O
(	O
MDP	O
)	O
,	O
which	O
,	O
in	O
RL	O
,	O
represents	O
the	O
problem	O
to	O
be	O
solved	O
.	O
</s>
<s>
The	O
transition	O
probability	O
distribution	O
(	O
or	O
transition	O
model	O
)	O
and	O
the	O
reward	O
function	O
are	O
often	O
collectively	O
called	O
the	O
"	O
model	O
"	O
of	O
the	O
environment	O
(	O
or	O
MDP	O
)	O
,	O
hence	O
the	O
name	O
"	O
model-free	B-Algorithm
"	O
.	O
</s>
<s>
A	O
model-free	B-Algorithm
RL	O
algorithm	O
can	O
be	O
thought	O
of	O
as	O
an	O
"	O
explicit	O
"	O
trial-and-error	O
algorithm	O
.	O
</s>
<s>
An	O
example	O
of	O
a	O
model-free	B-Algorithm
algorithm	O
is	O
Q-learning	B-Algorithm
.	O
</s>
