The usage of location information of weight vectors can help to overcome deยฎciencies of gradient-based learning for neural networks. We study the non-trivial structure of weight space, i.e. symmetries of feedforward networks in terms of their corresponding groups. We ยฎnd that these groups naturally
On Chu spaces in uncertainty analysis
โ Scribed by Hung T. Nguyen; Nhu T. Nguyen
- Publisher
- John Wiley and Sons
- Year
- 2000
- Tongue
- English
- Weight
- 139 KB
- Volume
- 15
- Category
- Article
- ISSN
- 0884-8173
No coin nor oath required. For personal study only.
โฆ Synopsis
A very general algebraic framework, known as Chu spaces, has recently been systematically used in modeling of information flow in distributed systems and in modeling of concurrency in computer science. In this paper, we explore this framework in the context of uncertainty modeling and analysis. We investigate two important Chu categories in probabilistic and fuzzy systems. We address fundamental questions in the modeling aspects toward applications.
๐ SIMILAR VOLUMES
A very promising method is proposed to solve the stability problem of steel space structures with system parametric uncertainties. Generalized random variables are presented to characterize such large complex space structures while reducing the number of variables in the structural analysis. The pro