Opened 16 years ago

Last modified 14 years ago

#51 new request

score object using SVM

Reported by: Peter Owned by: Peter
Priority: trivial Milestone: yat 0.x+
Component: classifier Version:
Keywords: Cc:

Description

By training an SVM and calculating the feature sensitivity one can rank the features based on that.

Change History (5)

comment:1 Changed 16 years ago by Peter

needs #32

comment:2 Changed 15 years ago by Peter

Milestone: later0.4

comment:3 Changed 14 years ago by Peter

Milestone: 0.40.5

comment:4 Changed 14 years ago by Peter

If we don't allow this to work polymorphically on KernelFunction, but limit us to the linear case, the SVM interface is sufficient to calculate the weight vector (which is the score we want)

That would imply that we don't need ticket:32

comment:5 Changed 14 years ago by Jari Häkkinen

Milestone: yat 0.5yat 0.x+
Note: See TracTickets for help on using tickets.