figure below reports the learning curves corresponding to classification accu$-$

racy and loss values. The values of these scores for both networks at conver$-$

gence are reported in the table. In the figure and table, the DNN implemented

in the code above is indicated with DNN$\_1,$ while the alternative DNN is indi$-$

cated with DNN $2.$

On the basis of the information in the figure and the table:

$($i$)$ Discuss the performance of the two DNNs$.$ What improvements $($if any$)$

could be made to improve the performance of the two models?

$($ii$)$ The network DNN$\_2$ has been implemented with additional regularization

compared to DNN$\_1.$ What type of regularization has been used? Support

your answer using the information from the figure.