We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Black box tests for algorithmic stability

Abstract: Algorithmic stability is a concept from learning theory that expresses the degree to which changes to the input data (e.g., removal of a single data point) may affect the outputs of a regression algorithm. Knowing an algorithm's stability properties is often useful for many downstream applications -- for example, stability is known to lead to desirable generalization properties and predictive inference guarantees. However, many modern algorithms currently used in practice are too complex for a theoretical analysis of their stability properties, and thus we can only attempt to establish these properties through an empirical exploration of the algorithm's behavior on various data sets. In this work, we lay out a formal statistical framework for this kind of "black box testing" without any assumptions on the algorithm or the data distribution, and establish fundamental bounds on the ability of any black box test to identify algorithmic stability.
Comments: 26 pages. Updates to Section 2.1.1 and Sections B.1 & B.2
Subjects: Machine Learning (cs.LG); Statistics Theory (math.ST)
Cite as: arXiv:2111.15546 [cs.LG]
  (or arXiv:2111.15546v2 [cs.LG] for this version)

Submission history

From: Byol Kim [view email]
[v1] Tue, 30 Nov 2021 16:36:58 GMT (21kb)
[v2] Wed, 25 May 2022 16:18:07 GMT (21kb)

Link back to: arXiv, form interface, contact.