Q$2)$

Apply the perceptron learning algorithm for the following pattern set until convergence or detection of a possible non$-$termination.

Recall that the perceptron learning equation is:

Start with weight vector: $(w0,w1,w2,w3{)}^T=(1,0,0,0{)}^T.$ Apply the patterns the patterns in same given order cyclically $(4$ steps in every iteration$).$

Stop when you have an iteration with each of the $4$ prediction evaluations are either True Positive or True Negative.

$(2,2,1{)}^T$inN

$(3,-3,5{)}^T$inP

$(2,0,-1{)}^T$inP

$(3,1,2{)}^T$inN

$($Remark: N means Negative $(0)$ and P means Positive $(1))$

Fill results in the following table. The first row of the first iteration is done for you. Note that each iteration is composed of $4$ rows.

You should add $4$ new rows for each additional iteration.