\item \points{25} {\bf Linear regression: linear in what?}

In the first two lectures, you have seen how to fit a linear function of the data for the regression problem.  In this question, we will see how linear regression can be used to fit non-linear functions of the data using feature maps. We will also explore some of its limitations, for which future lectures will discuss fixes.

\begin{enumerate}
	\input{featuremaps/01-degree-3-math}
        \ifnum\solutions=1 {
	  \input{featuremaps/01-degree-3-math-sol}
        }\fi

	\input{featuremaps/02-degree-3-code}
        \ifnum\solutions=1 {
	  \input{featuremaps/02-degree-3-code-sol}
        }\fi

	\input{featuremaps/03-degree-k-code}
        \ifnum\solutions=1 {
	  \input{featuremaps/03-degree-k-code-sol}
        }\fi

        \input{featuremaps/04-other-features}
        \ifnum\solutions=1 {
	  \input{featuremaps/04-other-features-sol}
        }\fi

        \input{featuremaps/05-overfitting}
        \ifnum\solutions=1 {
	  \input{featuremaps/05-overfitting-sol}
        }\fi
\end{enumerate}