Use two privacy protection methods $($k$-$anonymity and differential privacy$),$ for both training

and validation datasets. You need to choose five values for each method $($for example, k$=3,$

$5,9,12,$ and $15,$ and Epsilon$=0\mathrm{.}1,1,1\mathrm{.}25,1\mathrm{.}5,$ and $2).$ You may find the $$Score$$ and

$$Squared Error$$ values in the $$Analyze Utility $->$ Quality Model $->$ Properties$$ tab. Squared

Error is the measure of the percentage of error in the anonymized dataset. And Score talks

about the position of an observation above or below a distribution mean. You need to

generate four tables for two methods for two datasets as follows $($k$-$anonymity $$Table $1-$

training and Table $2-$Validation$,$ and differential privacy $$Table $1-$Training and Table $2-$

Validation$).$ From these tables, you need to report the best value for each method and each

dataset.