A Constructive Compound Neural Networks. II Application to Artificial Life in a Competitive Environment

We have developed a new efficient neural network-based algorithm for Alife application in a competitive world whereby the effects of interactions among organisms are evaluated in a weak form by exploiting the position of nearest food elements into consideration but not the positions of the other competing organisms. Two online learning algorithms, an instructive ASL (adaptive supervised learning) and an evaluative feedback-oriented RL (reinforcement learning) algorithm developed have been tested in simulating Alife environments with various neural network algorithms. The constructive compound neural network algorithm FuzGa [15] guided by the ASL learning algorithm has proved to be most efficient among the methods experimented including the classical constructive cascaded CasCor algorithm of [18], [19] and the fixed non-constructive fuzzy neural networks. Adopting an adaptively selected best sequence of feedback action period ∆α which we have found to be a decisive parameter in improving the network efficiency, the ASLguided FuzGa had a performance of an averaged fitness value of 541.8 (standard deviation 48.8) as compared with 500(53.8) for ASL-guided CasCor and 489.2 (39.7) for RL-guided FuzGa. Our FuzGa algorithm has also outperformed the CasCor in time complexity by 31.1%. We have elucidated how the dimensionless parameter food availability FA representing the intensity of interactions among the organisms relates to a best sequence of the feedback action period ∆α and an optimal number of hidden neurons for the given configuration of the networks. We confirm that the present solution successfully evaluates the effect of interactions at a larger FA, reducing to an isolated solution at a lower value of FA. The simulation is carried out by thread functions of Java by ensuring the randomness of individual activities. key words: neural networks construction, artificial life, fuzzy logic, genetic algorithm, reinforcement learning

[1]  Juichi Miyamichi,et al.  A New Constructive Compound Neural Networks Using Fuzzy Logic and Genetic Algorithm 1 Application to Artificial Life , 1998 .

[2]  Andrew W. Moore,et al.  Reinforcement Learning: A Survey , 1996, J. Artif. Intell. Res..

[3]  Frans C. A. Groen,et al.  The Optimal Number of Learning Samples and Hidden Units in Function Approximation With a Feedforward , 1993 .

[4]  D. Parisi,et al.  Phenotypic plasticity in evolving neural networks , 1994, Proceedings of PerAc '94. From Perception to Action.

[5]  Karl Sims,et al.  Evolving 3D Morphology and Behavior by Competition , 1994, Artificial Life.

[6]  D. Parisi,et al.  Growing neural networks , 1991 .

[7]  Liang Chen,et al.  A new approach for the automatic generation of membership functions and rules of multi-variable fuzzy system , 1995, Proceedings of ICNN'95 - International Conference on Neural Networks.

[8]  Kristian Lindgren,et al.  Evolutionary phenomena in simple dynamics , 1992 .

[9]  P. Anandan,et al.  Pattern-recognizing stochastic learning automata , 1985, IEEE Transactions on Systems, Man, and Cybernetics.

[10]  Christian Lebiere,et al.  The Cascade-Correlation Learning Architecture , 1989, NIPS.

[11]  Stefano Nolfi,et al.  Auto-teaching: networks that develop their own teaching input , 1993 .

[12]  Frank Dellaert,et al.  Toward an evolvable model of development for autonomous agent synthesis , 1994 .

[13]  Brad Fullmer and Risto Miikkulainen.  Using Marker-Based Genetic Encoding Of Neural Networks To Evolve Finite-State Behaviour , 1991 .

[14]  D. Yeung,et al.  Constructive feedforward neural networks for regression problems : a survey , 1995 .

[15]  David H. Ackley,et al.  Interactions between learning and evolution , 1991 .

[16]  Pawel Cichosz,et al.  Integrated Learning and Planning Based on Truncating Temporal Differences , 1997, ECML.