ЗАБЕЗПЕЧЕННЯ ТОЧНОСТІ ТА ПРОЗОРОСТІ НЕЧІТКОЇ МОДЕЛІ МАМДАНІ ПРИ НАВЧАННІ ЗА ЕКСПЕРИМЕНТАЛЬНИМИ ДАНИМИ
The typical violations of Mamdani-type fuzzy model that are produced while learning on experimental data, are described. The new learning scheme for Mamdani-type fuzzy model is proposed. The main features of that scheme are: 1) enlarged support of output variable fuzzy sets; 2) elimination of the co...
Збережено в:
| Дата: | 2007 |
|---|---|
| Автор: | |
| Формат: | Стаття |
| Мова: | Ukrainian |
| Опубліковано: |
V.M. Glushkov Institute of Cybernetics of NAS of Ukraine
2007
|
| Онлайн доступ: | https://jais.net.ua/index.php/files/article/view/331 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Problems of Control and Informatics |
Репозитарії
Problems of Control and Informatics| Резюме: | The typical violations of Mamdani-type fuzzy model that are produced while learning on experimental data, are described. The new learning scheme for Mamdani-type fuzzy model is proposed. The main features of that scheme are: 1) enlarged support of output variable fuzzy sets; 2) elimination of the cores of extreme fuzzy terms from tuning parameters; 3) insertion of constraint on linear order of fuzzy sets in frame of the term-set. The computational experiments show that the new learning scheme does not violate the fuzzy model transparency and at the same time the fuzzy model accuracy is no worse in comparison with typical learning. |
|---|