machine learning - how to do Classification based on the correlation of multiple features for a Supervised scenario -
i have 2 features: 'contact_last_name' , 'account_last_name' based on want classify data: logic if 2 features same i.e. contact_last_name same account_last_name - result 'success' or else 'denied'. so. example: if contact_last_name 'johnson' , account_last_name 'eigen' - result classified 'denied'. if both equal - 'edison' - result 'success'. how, can have classification algorithm set of data? [please note discard high correlation columns on here correlation between columns seems have logic classification]
i have tried use decision tree(c5.0) , naive bayes(naivebayes) in r both of these fail classify dataset correctly.
first of not use case machine learning, because can done string match, still if want give classification algorith, create table values 'contact_last_name' , 'account_last_name' , 'result' , give decision tree , predict third column.
note partition data training , testing.
Comments
Post a Comment