Difference between input attribute and predictable attribute

Question

Could anyone please clarify the difference between input attribute and predictable attribute for decision tree algorithm in Data mining.

Thanks.

Answer 1

These are concepts pertaining to Naive Bayesian models. Basically, an input attribute is an attribute that is given, i.e. something that you know as a fact from the outside, something you can observe. A predictable attribute is an attribute that cannot be observed directly but can be, hopefully, computed or somehow derived from a combination or relationship of various input attributes.

This is a decent explanation of a Naive Bayesian model implementation: http://technet.microsoft.com/en-us/library/ms174806.aspx

Hope it helps.