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

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

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

Statistics > Machine Learning

Title: Semi-analytical approximations to statistical moments of sigmoid and softmax mappings of normal variables

Abstract: This note is concerned with accurate and computationally efficient approximations of moments of Gaussian random variables passed through sigmoid or softmax mappings. These approximations are semi-analytical (i.e. they involve the numerical adjustment of parametric forms) and highly accurate (they yield 5% error at most). We also highlight a few niche applications of these approximations, which arise in the context of, e.g., drift-diffusion models of decision making or non-parametric data clustering approaches. We provide these as examples of efficient alternatives to more tedious derivations that would be needed if one was to approach the underlying mathematical issues in a more formal way. We hope that this technical note will be helpful to modellers facing similar mathematical issues, although maybe stemming from different academic prospects.
Subjects: Machine Learning (stat.ML); Neurons and Cognition (q-bio.NC)
Cite as: arXiv:1703.00091 [stat.ML]
  (or arXiv:1703.00091v2 [stat.ML] for this version)

Submission history

From: Jean Daunizeau [view email]
[v1] Wed, 1 Mar 2017 00:01:04 GMT (1265kb)
[v2] Fri, 3 Mar 2017 08:55:29 GMT (1307kb)

Link back to: arXiv, form interface, contact.