Section A

Q$1.$ Answer the following questions. $(5$ points each$)$

I. What are the different paradigms in Machine Learning? List any $2$ general tasks under each

paradigm.

II$.$ Describe a scenario when the model accuracy can be a misleading classification metric. How do

you deal with such a situation?

iil. Given some data $xin{R}^{nd}$ with $n$ samples and $d$ features, how to compute the covariance matrix?

IV$.$ You have collected a regression dataset with a total of $d$ features. After applying linear regression,

you obtained the set of optimal weights $\{w_0,w_1,w_2,w_{3}dots{w}_d\}.$ While keeping the output intact, if

we scale all the feature values by $2($multiply each dataset entry by $2$ ; excluding the output

values$),$ what will be the new optimal weights for this linear model?

V$.$ Discuss why the notion of distance fails in high$-$dimensional spaces.