D. Moriarty and R. Miikkulainen, Evolving Obstacle Avoidance Behavior in a Robot Arm. In From Animals to Animats: Proceedings of the Fourth International Conference on Simulation of Adaptative  Home/Search   Document Details and Download   Summary   Related Articles   Check

....population, the space of network specifications is searched for good teams of hidden units that result in high performance composite solutions. SANE has been applied successfully to several sequential decision tasks like the pole balancing problem [62] game playing [61] or robot arm control [63]. Another approach to cooperative coevolution is the one exploited by Paredis [72] In his work, a population of solutions and a population of permutations performed on the genotype of the first population coevolve. The motivation underlying this work is to limit the disruptive effect of search ....

David E. Moriarty and Risto Miikkulainen. Evolving obstacle avoidance behavior in a robot arm. In Pattie Maes, Maja J. Mataric, Jean-Arcady Meyer, Jordan Pollack, and Stewart W. Wilson, editors, Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior, pages 468--475, Cambridge, Massachusetts, 1996. MIT Press.

....subpopulation. 4. The Evaluation Recombination cycle is repeated until a network that performs suciently well in the task is found. Evolving networks at the neuron level has proven to be a very ecient method for solving reinforcement learning tasks such as pole balancing [6] robot arm control [8], and game playing [7] ESP is more ecient that SANE because the subpopulation architecture makes the evaluations more consistent in two ways: rst, the subpopulations that gradually form in SANE are already present by design in ESP. The species do not have to organize themselves out of a single ....

Moriarty, D.E., Miikkulainen, R.: Evolving obstacle avoidance behavior in a robot arm. Technical Report AI96-243, Department of Computer Sciences, The University of Texas at Austin (1996)

....that implement useful sub tasks, and a search for effective combinations of these sub tasks. SANE has been shown effective in several sequential decision tasks such as playing go (Richards et al. 1998) and Othello (Moriarty and Miikkulainen, 1995) and controlling a robot arm and a mobile robot (Moriarty and Miikkulainen, 1996). It was therefore selected as the starting point for the co evolution experiments reported in this paper. 3.3 COMPETITIVE CO EVOLUTION Co evolution is the simultaneous evolution of two or more populations with a common fitness landscape. Competitive co evolution can be defined as the evolution ....

Moriarty, D. and Miikkulainen, R. (1996). Evolving obstacle avoidance behavior in a robot arm. In Maes, P., Mataric, M., Meyer, J.-A., and Pollack, J., editors, From Animals to Animats: Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior, pages 468--475, Cambridge, MA. MIT Press.

No context found.

Moriarty, D. E., and Miikkulainen, R. (1996a). Evolving obstacle avoidance behavior in a robot arm. In From Animals to Animats: The Fourth International Conference on Simulation of Adaptive Behavior (SAB96).

....approaches. Populations were evolved for 80 generations as in the first experiment, but after each generation, the population diversity was measured. A diversity metric F can be generated by taking the average Hamming distance between every Evolutionary Computation Volume 5, Number 4 385 David E. Moriarty and Risto Miikkulainen 0 0.1 0.2 0.3 0.4 0.5 0 10 20 30 40 50 60 70 80 Diversity (Phi) Generation SANE Neuron SANE Standard Elite Standard Tournament Figure 7. The population diversity for each simulation. The neuron based approaches maintain very high levels of diversity, whereas the network based approaches converge to a ....

....a more explorative search. 5.2 Lesion Studies While the emergence of specializations is clear from the PCA studies, the function of each and the overall division of labor are not. To better understand the role of each specialization in Evolutionary Computation Volume 5, Number 4 389 David E. Moriarty and Risto Miikkulainen 30 20 10 0 10 20 30 30 20 10 0 10 20 30 Generation 0 30 20 10 0 10 20 30 30 20 10 0 10 20 30 Generation 10 30 20 10 0 10 20 30 30 20 10 0 10 20 30 Generation 20 30 20 10 0 10 20 30 30 20 10 0 10 20 30 Generation 40 30 20 10 0 10 20 30 30 20 10 0 10 20 ....

Moriarty, D. E., &Miikkulainen, R. (1996b). Evolving obstacle avoidance behavior in a robot arm. In From Animals to Animats: Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior (SAB-96) (pp. 468--475).

....realized, however, if the agent could separate the two blue states. Thus, any method that generates additional features to disambiguate states presents an important asset to EA 4. We used a binary tournament selection, a 50 policy population, 0.8 crossover probability, and 0.01 mutation rate. Moriarty, Schultz, Grefenstette 0 20 40 60 80 100 0 10 20 30 40 50 60 70 80 90 100 Generation Figure 14: The optimal policy distribution in the hidden state problem for an evolutionary algorithm. The graph plots the percentage of optimal policies in the population, averaged over 100 runs. methods. Kaelbling et al. 1996) describe ....

....those cases where the pole cannot be balanced, and concentrate on successful cases. This serves as another example of the advantages associated with search in policy space, based on overall policy performance, rather than paying too much attention to the value associated with individual states. Moriarty, Schultz, Grefenstette 10.4 Sane The Sane (Symbiotic, Adaptive Neuro Evolution) system was designed as a e cient method for building arti cial neural networks in RL domains where it is not possible to generate training data for normal supervised learning (Moriarty Miikkulainen, 1996a, 1998) The Sane system uses an ....

Moriarty, D. E., & Miikkulainen, R. (1996b). Evolving obstacle avoidance behavior in a robot arm. In From Animals to Animats: Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior (SAB-96), pp. 468-475 Cape Cod, MA.

No context found.

Moriarty, D. E., & Miikkulainen, R. (1996b). Evolving obstacle avoidance behavior in a robot arm. In From Animals to Animats: Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior (SAB-96), pp. 468--475 Cape Cod, MA.

....the game is so difficult that new techniques are probably going to be needed before go programs are as strong as those that play checkers, chess, or Othello. This paper explores the usefulness of neuro evolution as a mechanism for learning to play go. The SANE (Symbiotic, Adaptive Neuro Evolution [7, 8, 9]) algorithm demonstrates that networks that display a general ability in playing go on small boards can be evolved without To appear in Applied Intelligence. y This research was supported in part by NSF under grant #IRI 9504317. Sigma Xi Xi Xi Theta 2 3 Delta 1 Gamma Omega ....

....the many moves played were good and deserve credit for a win, and which were bad and deserve to be blamed for a loss. In go, this problem is severe enough that standard learning techniques such as backpropagation cannot be effectively applied. 5 SANE SANE 1 (Symbiotic Adaptive Neuro Evolution [7, 8, 9]) solves the credit assignment problem by using evolutionary algorithms to search for effective neural networks. Instead of punishing or rewarding individual moves, networks are evaluated, selected, and recombined based on their overall performance in the game. Evolutionary algorithms perform a ....

[Article contains additional citation context not shown here]

David E. Moriarty and Risto Miikkulainen. Evolving obstacle avoidance behavior in a robot arm. In P. Maes, M. Mataric, J.-A. Meyer, and J. Pollack, editors, From Animals to Animats: The Fourth International Conference on Simulation of Adaptive Behavior (SAB96), 1996.

No context found.

Moriarty, D. E., and Miikkulainen, R. (1996b). Evolving obstacle avoidance behavior in a robot arm.

No context found.

Moriarty, D. E., & Miikkulainen, R. (1996b). Evolving obstacle avoidance behavior in a robot arm. In From Animals to Animats: Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior (SAB-96), pp. 468--475 Cape Cod, MA.

No context found.

David E. Moriarty and Risto Miikkulainen. Evolving obstacle avoidance behavior in a robot arm. In P. Maes, M. Mataric, J.-A. Meyer, and J. Pollack, editors, From Animals to Animats: The Fourth International Conference on Simulation of Adaptive Behavior (SAB96), 1996.

....from game playing to robotics require memory to disambiguate states. The current task, therefore, can serve as a surrogate with which new methods can be tested. SANE has been shown effective in a variety of domains, including robot arm control, constraint satisfaction, and in controlling chaos [6, 8, 9]. The results presented in this paper show that ESP extends this powerful method by allowing the evolution of recurrent networks and therefore making it applicable to non Markovian environments. In the future, we plan to apply this system to real world tasks such as robot navigation and game ....

D. E. Moriarty and R. Miikkulainen. Evolving obstacle avoidance behavior in a robot arm. In P. Maes, M. Mataric, J.-A. Meyer, and J. Pollack, editors, From Animals to Animats 4: Proceedings of the 4th International Conference on Simulation of Adaptive Behavior, pages 468--475, Cambridge, MA, 1996. MIT Press.

No context found.

Moriarty, D. E., & Miikkulainen, R. (1996b). Evolving obstacle avoidance behavior in a robot arm. In From Animals to Animats: Proceedings of the Fourth International Conference on Simulation of Adaptive Behavior (SAB-96), pp. 468--475 Cape Cod, MA.

No context found.

Moriarty, D. E., and Miikkulainen, R. (1996b). Evolving obstacle avoidance behavior in a robot arm. In Maes, P., Mataric, M., Meyer, J.-A., and Pollack, J., editors, From Animals to Animats: The Fourth International Conference on Simulation of Adaptive Behavior (SAB96).

No context found.

D. Moriarty and R. Miikkulainen, Evolving Obstacle Avoidance Behavior in a Robot Arm. In From Animals to Animats: Proceedings of the Fourth International Conference on Simulation of Adaptative

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC