Name
Abstract | ||
|---|---|---|
It is investigated in this paper that how learning algorithm of fuzzy system can be arranged by gradient descent method and how the learning speed can be increased in this method. First, the optimal range of learning speed coefficient not to be trapped in local minima and not to provide too slow learning speed is investigated. With the optimal range of learning speed coefficient, the optimal value of learning speed coefficient is suggested. With this value, the learning algorithm should not give learning oscillations and not provide too slow learning speed in any system to be approximated. Modified momentum is developed and applied to the learning scheme of gradient descent method in order to increase the learning speed. Simulation results assure that this modified momentum provides fast learning speed and also can converge to the optimal point within stable learning process without selecting the momentum coefficient arbitrarily. |
Year | DOI | Venue |
|---|---|---|
1996 | 10.1016/0165-0114(95)00081-X | Fuzzy Sets and Systems |
Keywords | Field | DocType |
gradient descent method,local minima,fuzzy system,oscillations,gradient descent | Online machine learning,Stochastic gradient descent,Gradient descent,Control theory,Wake-sleep algorithm,Maxima and minima,Artificial intelligence,Momentum,Fuzzy control system,Feature learning,Mathematics,Machine learning | Journal |
Volume | Issue | ISSN |
77 | 3 | 0165-0114 |
Citations | PageRank | References |
2 | 0.49 | 1 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Gee Yong Park | 1 | 2 | 0.49 |
| Poong-hyun Seong | 2 | 115 | 24.53 |