 | Witold Pedrycz, James F. Peters - 1998 - 504 lapas
...scheme. For the same reason (ie software engineering expert knowledge is available) we strongly oppose using learning strategies that are only result-driven...we see the necessity of such approaches when only few guiding principles are available and sufficient project data can be utilized for supervised learning.... | |
 | Miriam A. Drake - 2003 - 874 lapas
...scheme. For the same reason (ie, software engineering expert knowledge is available) we strongly oppose using learning strategies that are only result-driven...project data can be utilized for supervised learning. CONCLUSIONS The choice of the proper approach to automatic decision support depends on the problem.... | |
 | Miriam A. Drake - 2003 - 874 lapas
...strategies that are only result-driven 1e.g., classification trees or mere neural network approaches1. However, we see the necessity of such approaches when...project data can be utilized for supervised learning. CONCLUSIONS The choice of the proper approach to automatic decision support depends on the problem.... | |
| |
Since there are some guiding principles for decision support available, we focus on utilizing expert-derived, however vague, knowledge that we included in a fuzzy expert system-type classification scheme. If software engineering expert knowledge is available... 