# How to apply ANOVA test to K-Means clustering

I'm generally looking for how to apply ANOVA test to neural networks.I'm particularly interested in ANOVA tests for

• clustering
• function estimation

below is the code taken from K-means clustering from k-means

x = {{1, 2, 3, 4, 5}, {6, 7, 8, 9, 10}, {11, 12, 13, 14,
15}, {16, 17, 18, 19, 20}, {21, 22, 23, 24, 25}};
cc = ClusteringComponents[x, 2, 1, Method -> "KMeans",
"DistanceFunction" -> SquaredEuclideanDistance, "RandomSeed" -> 1]


The question is how to apply ANOVA[] tests in mathematica to clustering and function estimation?

-
what have you tried? Have you read the docs for Anova[]? Sounds like your question is more on the statistical side of things... –  rm -rf Jul 8 '12 at 14:41
Yes. I know 'Anova[]' also for fitting ... But my Q is more specific to ANN. also not directly subject of statistics. –  s.s.o Jul 8 '12 at 14:54
I don't think this question is on-topic here. It's not about how to use Mathematica's ANOVA function, but how to use an ANOVA to analyze a given clustering. The more appropriate site for that type of questions is stats.stackexchange.com –  Sjoerd C. de Vries Jul 8 '12 at 15:15
It's not really about analyzing a cluster. But to apply Mathematica ANOVA function to a NN fitting problem. You can get ANOVA for linear or non linear fitting in Mathematica. But there is no example or doc to this sort of fitting. –  s.s.o Jul 8 '12 at 15:22
Mathematica provides ANOVA tables after a NonlinearModelFit. Plenty of examples on that in the documentation. There's also the ANOVA package. Did you have a look at that? If that didn't help I guess you have to clarify your question. It is not clear at all what you want to do. –  Sjoerd C. de Vries Jul 8 '12 at 17:04
show 1 more comment