Filters








36,107 Hits in 4.9 sec

Developing hierarchical anticipations via neural network-based event segmentation [article]

Christian Gumbsch, Maurits Adam, Birgit Elsner, Georg Martius, Martin V.Butz
2022 arXiv   pre-print
Using a simulated robotic manipulator, we demonstrate that the system (i) learns latent states that accurately reflect the event structure of the data, (ii) develops meaningful temporal abstract predictions  ...  A higher level network learns to predict the situations in which the latent states tend to change.  ...  Hypothesis for modeling infants' goal anticipations. (a): Infants first learn event representations (top).  ... 
arXiv:2206.02042v1 fatcat:jgwlsvaab5aojhmllpgvjqnhcq

Internalizing Knowledge for Anticipatory Classifier Systems in Discretized Real-Valued Environments

Norbert Kozlowski, Olgierd Unold
2022 IEEE Access  
Despite increased input size, all possible environmental transitions were learned latently, without obtaining any explicit incentives.  ...  This work demonstrates that Anticipatory Learning Classifier Systems (ALCS) can successfully build sets of conditional rules foreseeing the consequences of executed actions.  ...  ACS AND ACS2 Hoffmann proposed a theory of a psychological theory of anticipatory behavioural control [30] , stating that conditional action-effects relations are learned latently using anticipations  ... 
doi:10.1109/access.2022.3162925 fatcat:dvmpcp4rr5axthijlus6htx4zq

Goal-directed Planning and Goal Understanding by Active Inference: Evaluation Through Simulated and Physical Robot Experiments [article]

Takazumi Matsumoto, Wataru Ohata, Fabien C. Y. Benureau, Jun Tani
2022 arXiv   pre-print
We show that goal-directed action planning and generation in a teleological framework can be formulated using the free energy principle.  ...  goal-directed action plans, but can also understand goals by sensory observation, and (3) the model generates future action plans for given goals based on the best estimate of the current state, inferred  ...  This allows agents using the T-GLean model to anticipate goals and future actions by observing the actions of another agent.  ... 
arXiv:2202.09976v1 fatcat:om6waelch5fhpb3rw62gl6tm5q

Goal-Directed Planning and Goal Understanding by Extended Active Inference: Evaluation through Simulated and Physical Robot Experiments

Takazumi Matsumoto, Wataru Ohata, Fabien C. Y. Benureau, Jun Tani
2022 Entropy  
goal-directed action plans, but can also understand goals through sensory observation, and (3) the model generates future action plans for given goals based on the best estimate of the current state,  ...  We show that goal-directed action planning and generation in a teleological framework can be formulated by extending the active inference framework.  ...  This allows agents using the T-GLean model to anticipate goals and future actions by observing the actions of another agent.  ... 
doi:10.3390/e24040469 pmid:35455132 pmcid:PMC9026632 fatcat:jefxqguvfndslffgvvpxiij2q4

Shared Autonomy with Learned Latent Actions [article]

Hong Jun Jeon, Dylan P. Losey, Dorsa Sadigh
2020 arXiv   pre-print
In this work, we adopt learned latent actions for shared autonomy by proposing a new model structure that changes the meaning of the human's input based on the robot's confidence of the goal.  ...  Our insight is that---by combining intuitive embeddings from learned latent actions with robotic assistance from shared autonomy---we can enable precise assistive manipulation.  ...  We test latent actions used by themselves (LA), as well as latent actions combined with shared autonomy (LA+SA).  ... 
arXiv:2005.03210v2 fatcat:tytd4w46t5hrrdcniifskr5ffq

Learning Classifier Systems [chapter]

Tim Kovacs
2004 Strength or Accuracy: Credit Assignment in Learning Classifier Systems  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1007/978-0-85729-416-6_2 fatcat:zen5oko3qrg5dce3bxzsbhgope

Learning Classifier Systems [chapter]

Geoffrey I. Webb, Claude Sammut, Claudia Perlich, Tamás Horváth, Stefan Wrobel, Kevin B. Korb, William Stafford Noble, Christina Leslie, Michail G. Lagoudakis, Novi Quadrianto, Wray L. Buntine, Novi Quadrianto (+9 others)
2011 Encyclopedia of Machine Learning  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1007/978-0-387-30164-8_449 fatcat:z6lgoifyi5ephapyrrrdjoa2pq

Learning Classifier Systems [chapter]

2008 Using Artificial Intelligence in Chemistry and Biology  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1201/9780849384141.ch9 fatcat:6zbqx7caynh4nfatqjqzyt7omy

Learning classifier systems

Martin V. Butz
2011 Proceedings of the 13th annual conference companion on Genetic and evolutionary computation - GECCO '11  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1145/2001858.2002121 dblp:conf/gecco/Butz11 fatcat:jozv6y67ufbolno63hm5mfxi2i

Learning classifier systems

Martin V. Butz
2007 Proceedings of the 2007 GECCO conference companion on Genetic and evolutionary computation - GECCO '07  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1145/1274000.1274104 dblp:conf/gecco/Butz07 fatcat:sdj2g52wfjbqbp6wmmel7c3ore

Learning classifier systems

Pier Luca Lanzi
2009 Proceedings of the 11th annual conference companion on Genetic and evolutionary computation conference - GECCO '09  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1145/1570256.1570406 dblp:conf/gecco/Lanzi09 fatcat:pcsno5nxp5f5neqit5qvcxluie

Learning classifier systems

Will N. Browne, Ryan Urbanowicz
2013 Proceeding of the fifteenth annual conference companion on Genetic and evolutionary computation conference companion - GECCO '13 Companion  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1145/2464576.2483909 dblp:conf/gecco/BrowneU13 fatcat:imns7hwz7vbm5lzny6je26t3n4

Learning classifier systems

Martin V. Butz
2010 Proceedings of the 12th annual conference comp on Genetic and evolutionary computation - GECCO '10  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1145/1830761.1830898 dblp:conf/gecco/Butz10 fatcat:ytrlwhonnnetdjitafvwlcb2se

Learning Classifier Systems

Larry Bull, Pler Luca Lanzi, Wolfgang Stolzmann
2002 Soft Computing - A Fusion of Foundations, Methodologies and Applications  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1007/s005000100110 fatcat:dxn5lz4o6vgh3olkzx4azjlw2e

Learning Classifier Systems [chapter]

A. E. Eiben, J. E. Smith
2003 Natural Computing Series  
That is, ZCS is shown to learn under a latent learning scenario using the lookahead scheme. Its ability to form maps in reinforcement learning tasks is then considered.  ...  Learning Classifier Systems use reinforcement learning, evolutionary computing and/or heuristics to develop adaptive systems.  ...  ACKNOWLEDGEMENT Thanks to the members of the Learning Classifier System Group at UWE for many useful discussions during this work.  ... 
doi:10.1007/978-3-662-05094-1_7 fatcat:pyc25gpi5vgvdjgsvzo2cbnxwa
« Previous Showing results 1 — 15 out of 36,107 results