Given the training dataset with negative samples x1 =
(3, 3)T , x2 = (4, 3)T and positive
samples x3 = (1, 1)T , try using primal form
of perceptron learning algorithm to get the per- ceptron
model: f (x)
= sign(w · x + b) ,
where w = (w(1), w(2))T , x = (x(1),
x(2))T . Let the initialvaluesofwandbbew0 = 0andb0 =
0,thelearningrate? = 1,andthesequenceof misclassified samples
be x1, x3, x3, x3, x1, x3, x3.
Calculate the corresponding w and b of each
misclassified sample.