Question $6[1\mathrm{.}5$ pts$]$ Assuming we have two sets of instances, which belong to two classes, with Mean squared errors $E(W)$ after the first$-$round weight updating:

Second Round

$\backslash$table$[[$Input$,$Weight,v$,$Desired,Output,$\backslash$Delta w$],[(1,1,0),,,,,],[(1,-1,0),,,,,],[(1,0,-1),,,,,],[(1,0,1),,,,,],[(1,1,1),,,,,],[(1,-1,-1),,,,,]]$

New weight after second round:

Mean squared errors $E(W)$ after the second$-$round weight updating:

each class containing three instances. $C1=\{(1,0),(1,1),(0,-1)\}$;$C2=\{(0,1),(-1,0),(-1,-1)\}.$

Assuming the class label for $C1$ and $C2$ are $1$ and $-1,$ respectively, the learning $=0\mathrm{.}1,$ and the

initial weights are $w_0=1,w_1=1,$ and $w_2=1.$ Please use gradient learning rule to learn a linear

decision surface to separate the two classes. List the results in the first two rounds by using

tables in the following form $($Report the mean squared errors of all instances with respect to

the initial weight values, and the mean squared errors E$($W$)$ AFTER the weight updating for each

round$).$

Mean squared errors $E(W)$ corresponding to the initial weights:

First Round

New weight after first round: