machine learning - How WEKA compute Sum of Squared-Error Value or SSE? -
i new weka.
i know how weka sse value of simple k-means algorithm?
my friend , implemented java implemented k-means algorithm, , same dataset, our java implemented algorithm sse value of around 400 while weka around 2000. how possible?
my friend said weka uses k-means++. 1 of reasons make them have different result?
any idea appreciated. , please correct me if there's wrong. love learn.
have normalized data?
different normalization cause both different results , different sse values.
also try exporting result, , using same implementation compute both sse values.
Comments
Post a Comment