Abstract:
A general technique is proposed for embedding on-line clustering algorithms based on competitive learning in a reinforcement learning framework. The basic idea is that the clustering system can be viewed as a reinforcement learning system that learns through reinforcements to follow the clustering strategy we wish to implement. In this sense, the RGCL (Reinforcement Guided Competitive Learning) algorithm is proposed that constitutes a reinforcement-based adaptation of LVQ with enhanced clustering capabilities. In addition, we suggest extensions of RGCL and LVQ that are characterized by the property of sustained exploration and significantly improve the performance of those algorithms as indicated by experimental tests on well-known datasets. 1
Citations
|
887
|
Reinforcement learning: A survey
– Kaelbling, Littman, et al.
- 1996
|
|
718
|
Pattern recognition and neural networks
– Ripley
- 1996
|
|
252
|
A stochastic approximation method
– Robbins, Munro
- 1951
|
|
201
|
Self-Organization and Associative
– Kohonen
- 1990
|
|
200
|
A Connectionist Machine for Genetic Hillclimbing
– Ackley
- 1987
|
|
174
|
Simple statistical gradient-following algorithms for connectionist reinforcement learning
– Williams
- 1992
|
|
172
|
Neural-gas Network for Vector Quantization and Its Application to Time-series Prediction
– Martinetz, Berkovich, et al.
- 1993
|
|
98
|
Competitive learning algorithms for vector quantization
– Ahalt, Krishnamurthy, et al.
- 1990
|
|
60
|
The Irises of the Gaspe Peninsula
– Anderson
- 1935
|
|
52
|
Pattern–recognizing stochastic learning automata’, in
– Barto, Anandan
- 1985
|
|
22
|
Optimization of Clustering Criteria by Reformulation
– Hathaway, Bezdek
- 1995
|
|
7
|
Toward a theory of reinforcement learning connectionist systems
– Williams
- 1988
|
|
1
|
Enhancing Stochasticity in Reinforcement Learning Schemes: Application to the Exploration of Binary Domains
– Kontoravdis, Likas, et al.
- 1995
|
|
1
|
Function Optimization Using Connectionnist Reinforcement Learning Networks
– Williams, Peng
- 1991
|
|
