Abstract
Neural networks are beginning to be used for modeling of complex systems, often in process and quality control of manufacturing processes. Since neural networks are wholly empirical models, the resulting model is highly dependent on the data used to construct it. Using data from production lines ensures integrity, however may produce biased samples which do not reflect the entire range of domain operation. Supplementing production observations with data gathered from designed experiments alleviates this problem of overly focused data sets. This paper describes this designed experiment supplemental approach as it was used on two research projects involving complex manufacturing processes.
Original language | American English |
---|---|
Pages | 229-238 |
Number of pages | 10 |
State | Published - 1995 |
Externally published | Yes |
Event | Proceedings of the 1995 4th Industrial Engineering Research Conference - Nashville, TN, USA Duration: May 24 1995 → May 25 1995 |
Other
Other | Proceedings of the 1995 4th Industrial Engineering Research Conference |
---|---|
City | Nashville, TN, USA |
Period | 5/24/95 → 5/25/95 |
ASJC Scopus subject areas
- Industrial and Manufacturing Engineering