A Different Type of Convergence for Statistical Learning Algorithms

@inproceedings{Rudin2003ADT,
  title={A Different Type of Convergence for Statistical Learning Algorithms},
  author={Cynthia Rudin},
  year={2003}
}
We discuss stability for a class of learning algorithms with respect to noisy labels. The algorithms we consider are for regression, and they involve the minimization of regularized risk functionals, such as L(f) := 1 N PN i=1(f(xi)− yi) +λ‖f‖H. We shall call the algorithm ‘stable’ if, when yi is a noisy version of f(xi) for some functionf ∈ H, the output of the algorithm converges to f as the regularization term and noise simultaneously vanish. We consider two flavors of this problem, one… CONTINUE READING

References

Publications referenced by this paper.
Showing 1-10 of 10 references

Methods of Modern Mathematical Physics I: Functional Analysis

  • M. Reed, B. Simon
  • 1980
1 Excerpt

Solution of Ill-Posed Problems

  • A. N. Tikhonov, V.Y.Arsenin
  • 1977
1 Excerpt

Similar Papers

Loading similar papers…