Future Work

We will analyze the generality, in terms of different environments and starting points, of the parameters obtained by the genetic algorithm. Further work should also focus on designing an agent capable of identifying the complexity of the task being performed, so that the parameters can be switched from one set to another. We will explore the use of Case Base Reasoning techniques on this ``situation identifier'' agent.

Table 5.3: Results obtained by the different parameter sets

	Going to $C_1$			Going to $C_2$
	$\overline{s}$	$\overline{c}$	$f$	$\overline{s}$	$\overline{c}$	$f$
$C_1$ set	1	50.5	0.017	0.5	127.5	0.003
$C_2$ set	0.5	42.5	0.011	1	122	0.007
HT set	0.5	69	0.005	0	-	0

Figure 5.16: Going to cluster $C_1$

Figure 5.17: Going to cluster $C_2$