Spine yes are training nural Norterk with a Backpropagation algorithm on two diffe Following are the two different plots obtained for these te different tasks. With respect tis the given plots arwwer the following questions $[4-2-6$ Marks$]$

$($A$)$

$0\mathrm{.}000$

$0\mathrm{.}00\%$

ANI

:

Training set crue Validation set er

$10\mathrm{.}00\%$

$0\mathrm{.}005$

$0\mathrm{.}004$

$0\mathrm{.}003$

$0\mathrm{.}000$

Number of weight updates

al Comment on the representativeness of the training dataset, Justify your answer.

When number of iterations reaches approximately $9000,$ training should be stopped or to be allowed to continue? Justify your answer.

Suppose no$.$ of iterations is set as the convergence criteria, and you are required to experiment with two cases as follows. Comment on the corectness of the decision in both the cases. Give justification

Case I when training is stopped at $4000$ iterations and learned models acce $||$ Case it when training is stopped at $20000$ iterations and leamed model is not