BROWSE
Detail
Conference Paper
The analysis of a stochastic differential approach for Langevin competitive learning algorithm
Cited - time in 
Share 
Share
- Authors
- Jinwuk Seok, Jeun-Woo Lee
- Issue Date
- 2002-08
- Citation
- International Conference on Pattern Recognition (ICPR) 2002, pp.80-83
- Language
- English
- Type
- Conference Paper
- Abstract
- Recently, various types of neural network models have been used successfully for applications in pattern recognition, control, signal processing, and so on. However, the previous models are not suitable for hardware implementation due to their complexity. In this paper, we present a survey of stochastic analysis for Langevine Competitive Learning Algorithm, known that it is easy for hardware implementation [1]. Since the Langevine competitive learning algorithm uses a time-invariant learning rate and a stochastic reinforcement term, it is necessary to analyze with stochastic differential or difference equation. The result of analysis verifies that the Langevine Competitive Learning process is equal to the standard Ornstein-Uhlenbeck process and has weak convergence property. The experimental results for Gaussian distributed data shows that the analysis provided in this paper is available.
- KSP Keywords
- Competitive learning algorithm, Convergence properties, Differential approach, Hardware Implementation, Learning rate, Ornstein-Uhlenbeck process, Pattern recognition, Signal Processing, Stochastic differential, Time-invariant, difference equation