Hello list, I'm considering using Mahout for some my my classification work. I couldn't find any references to how the Bayesian classifier treats its data so I thought I'd ask here. I've currently got my own routines in Mathematica.
My problem is I have very non-guassian distributions - bimodal or multi-modal. As an example think classifying employed vs unemployed. The employed would clump together between 18 and 60, were as the unemployed would be at the far extremes under 18 and over 60. Can mahout deal with these sorts of distributions in data? Cheers
