Class Prediction by Prediction Intervals for Neural Nets
Abstract:
Generally, the unknown coefficients of neural nets are estimated by nonlinear least squares. Therefore, prediction intervals for the true value of the target feature exist. The paper proposes to use such intervals for class prediction and model selection. Only in this way, the uncertainty of class predictions can be indicated.
| Zugehörige Institution(en) am KIT | Institut für Wirtschaftsinformatik und Marketing (IISM) |
| Publikationstyp | Zeitschriftenaufsatz |
| Publikationsjahr | 2020 |
| Sprache | Englisch |
| Identifikator | ISSN: 2363-9881 KITopen-ID: 1000125696 |
| Erschienen in | Archives of Data Science, Series A |
| Band | 6 |
| Heft | 1 |
| Seiten | P03, 13 S. online |