\documentclass[fleqn]{article}
\usepackage{mydefs}
\usepackage{notes}
\usepackage{url}
\begin{document}
\lecture{Machine Learning}{HW02: Nearest neighbor classifier}{CS 689, Spring 2015}
% IF YOU ARE USING THIS .TEX FILE AS A TEMPLATE, PLEASE REPLACE
% "CS 689, Spring 2015" WITH YOUR NAME AND UID.
Hand in via moodle at: \url{https://moodle.umass.edu/course/view.php?id=20836}.
Remember that only PDF submissions are accepted. We encourage using
\LaTeX\ to produce your writeups. See \verb+hw00.tex+ for an example
of how to do so. You can make a \verb+.pdf+ out of the \verb+.tex+ by
running ``\verb+pdflatex hw00.tex+''. You'll need mydefs.sty and notes.sty which can be downloaded from the course page.
\bee
\i Give an example of a low dimensional (approx. 20 dimensions), medium dimensional (approx. 1000 dimensions) and high dimensional (approx. 100000 dimensions) problem that you care about.
% \begin{solution}
%\end{solution}
\i What does the decision boundary of 1 nearest neighbor classifier for 2 points (one positive, one negative) look like?
% \begin{solution}
% \end{solution}
\i Clustering was introduced as a way to speed up k nearest neighbor (kNN) classification. Is it possible that clustering can lead to a better classifier? Briefly explain why.
% \begin{solution}
%\end{solution}
\i Give two examples of data where the Euclidean distance is not the right metric.
\i Does the accuracy of a kNN classifier using the Euclidean distance change if you (a) translate the data (b) scale the data (i.e., multiply the all the points by a constant), or (c) rotate the data? Explain. Answer the same for a kNN classifier using Manhattan distance\footnote{\url{http://en.wikipedia.org/wiki/Taxicab_geometry}}.
\ene
\end{document}