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: Loss convergence in a causal Bayesian neural network of retail firm performance

Abstract: We extend the empirical results from the structural equation model (SEM) published in the paper Assortment Planning for Retail Buying, Retail Store Operations, and Firm Performance [1] by implementing the directed acyclic graph as a causal Bayesian neural network. Neural network convergence is shown to improve with the removal of the node with the weakest SEM path when variational inference is provided by perturbing weights with Flipout layers, while results from perturbing weights at the output with the Vadam optimizer are inconclusive.
Comments: 7 pages, 5 figures, 2 tables
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG)
ACM classes: I.5.1
Cite as: arXiv:2008.13038 [stat.ML]
  (or arXiv:2008.13038v1 [stat.ML] for this version)

Submission history

From: F. Trevor Rogers [view email]
[v1] Sat, 29 Aug 2020 19:16:43 GMT (766kb)

Link back to: arXiv, form interface, contact.