this is for holding javascript data
JP Breuer edited subsubsection_Comparison_of_the_GMM__.tex
about 8 years ago
Commit id: 7fbf183614273a7f8715b9e92ad0168978a4711d
deletions | additions
diff --git a/subsubsection_Comparison_of_the_GMM__.tex b/subsubsection_Comparison_of_the_GMM__.tex
index c8852c1..8ad678e 100644
--- a/subsubsection_Comparison_of_the_GMM__.tex
+++ b/subsubsection_Comparison_of_the_GMM__.tex
...
\subsection{Comparison of the GMM and Hiearchical Clustering methods}
Comparing the two clustering methods described in the previous section, it is possible to see very different results when looking at the first two principal components plotted against each other. For the Gaussian Mixture Model, it is possible to see the corresponding clusters for Asian, European, and American cars, as well as their specific centroids and covariance ellipses. Based solely on this plot, it is quite difficult to make any immediately confirmable conclusions regarding any of the three classes due to a large overlap in their covariance matrices. Naturally, the clustering for the observations outside of these regions of interest can be more clearly identified and separated into their own clusters without much possibility of misclustering.