Hi,
I'm trying to make a comprehensive analysis of feature significance for a Machine Learning scheme of mine and while I was trying to parallelize the process I stumbled upon a post mentioning that the dev edition supports parallelism.
I am currently using v3.6.13 and I wanted to run a 10-fold cross validation exhaustive search for Consistency Subset Eval feature selection. It's been running for days now and I'm only in the 8th fold so parallel is the way to go.
Upon launching the dev version I noticed that there is no more Consistency Subset Eval neither the option for exhaustive search. Am I missing something? Are certain algorithms exclude from the 3.7 edition of WEKA as obsolete or inefficient?
Thanks.
I'm trying to make a comprehensive analysis of feature significance for a Machine Learning scheme of mine and while I was trying to parallelize the process I stumbled upon a post mentioning that the dev edition supports parallelism.
I am currently using v3.6.13 and I wanted to run a 10-fold cross validation exhaustive search for Consistency Subset Eval feature selection. It's been running for days now and I'm only in the 8th fold so parallel is the way to go.
Upon launching the dev version I noticed that there is no more Consistency Subset Eval neither the option for exhaustive search. Am I missing something? Are certain algorithms exclude from the 3.7 edition of WEKA as obsolete or inefficient?
Thanks.