On the Stability of Empirical Risk Minimization in the Presence of Multiple Risk Minimizers

Details On the Stability of Empirical Risk Minimization in the Presence of Multiple Risk Minimizers On the Stability of Empirical Risk Minimization in the Presence of Multiple Risk Minimizers

2010-02-10

arxiv.org/abs/1002.2044v1

Computer Science

Learning

4 pages

Scientific paper

Recently Kutin and Niyogi investigated several notions of algorithmic stability--a property of a learning map conceptually similar to continuity--showing that training-stability is sufficient for consistency of Empirical Risk Minimization while distribution-free CV-stability is necessary and sufficient for having finite VC-dimension. This paper concerns a phase transition in the training stability of ERM, conjectured by the same authors. Kutin and Niyogi proved that ERM on finite hypothesis spaces containing a unique risk minimizer has training stability that scales exponentially with sample size, and conjectured that the existence of multiple risk minimizers prevents even super-quadratic convergence. We prove this result for the strictly weaker notion of CV-stability, positively resolving the conjecture.

Affiliated with

Also associated with

No associations

Rating

On the Stability of Empirical Risk Minimization in the Presence of Multiple Risk Minimizers does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.

If you have personal experience with On the Stability of Empirical Risk Minimization in the Presence of Multiple Risk Minimizers, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and On the Stability of Empirical Risk Minimization in the Presence of Multiple Risk Minimizers will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFWR-SCP-O-238522