𝔖 Bobbio Scriptorium
✦   LIBER   ✦

Feedforward neural network's sensitivity to input data representation

✍ Scribed by Igor T. Podolak

Publisher
Elsevier Science
Year
1999
Tongue
English
Weight
434 KB
Volume
117
Category
Article
ISSN
0010-4655

No coin nor oath required. For personal study only.

✦ Synopsis

Neural networks can be used to develop solutions to problems which are strictly symbolic. A question arises how to represent symbols in terms of number vectors understandable to neural networks. Data representation used should promote good generalization and reduce simulation uncertainty of the resulting model. Straightforward methods, which are most widely used, result in large networks which can prohibit solution of large problems. In the paper some new methods, which try to build information about the problem at hand into the representation, are proposed. It is shown that they are less sensitive to input data errors. @ 1999 Elsevier Science B.V.

📜 SIMILAR VOLUMES

Using feedforward neural networks and fo
Using feedforward neural networks and forward selection of input variables for an ergonomics data classification problem
✍ Chuen-Lung Chen; David B. Kaber; Patrick G. Dempsey 📂 Article 📅 2003 🏛 John Wiley and Sons 🌐 English ⚖ 829 KB

## Abstract A method was developed to accurately predict the risk of injuries in industrial jobs based on datasets not meeting the assumptions of parametric statistical tools, or being incomplete. Previous research used a backward‐elimination process for feedforward neural network (FNN) input varia

Sensitivity analysis of availability est
Sensitivity analysis of availability estimates to input data characterization using design of experiments
✍ Darren P. Durkee; Edward A. Pohl; Edward F. Mykytka 📂 Article 📅 1998 🏛 John Wiley and Sons 🌐 English ⚖ 102 KB 👁 2 views

Reliability analysts are often faced with the challenge of characterizing the behaviour of system components based on limited data. The purpose of this study is to provide insight into which availability model input data are most significant and how many data are necessary to achieve desired accurac

Sensitivity of mass transfer-based evapo
Sensitivity of mass transfer-based evaporation equations to errors in daily and monthly input data
✍ V. P. SINGH; C.-Y. XU 📂 Article 📅 1997 🏛 John Wiley and Sons 🌐 English ⚖ 152 KB

The in¯uence of data errors on the performance of mass transfer-based evaporation equations was investigated for both monthly and daily data from a climatological station in the State of Vaud in Switzerland. Evaporation estimates were found to be particularly sensitive to vapour pressure gradient, l