Reinforcement Learning Module


 


Integrating temporal difference methods and self-organizing neural networks for reinforcement learning with delayed evaluative feedback AH Tan, N Lu, D Xiao – Neural Networks, IEEE Transactions on, 2008 – ieeexplore.ieee.org … The clusters are then used as the compressed states and actions by a separate Q-learning module. Another line of work by Ninomiya [33] couples a supervised ART system with a TD reinforcement learning module in a hybrid architecture. … Cited by 46 Related articles BL Direct All 10 versions

[PDF] from robotcub.org Reinforcement learning for imitating constrained reaching movements F Guenter, M Hersch, S Calinon, A Billard – Advanced Robotics, 2007 – ingentaconnect.com … Our system is based on a dynamic system generator modulated by a learned speed trajectory. This system is combined with a reinforcement learning module to allow the robot to adapt the trajectory when facing a new situation, eg, in the presence of obstacles. … Cited by 54 Related articles All 20 versions

[PDF] from tamu.edu [PDF] Reinforcement learning of morphing airfoils with aerodynamic and structural effects A Lampton, A Niksch, J Valasek – Journal of Aerospace …, 2009 – jungfrau.tamu.edu … This paper first describes the reinforcement learning module. It uses Q-learning to learn how to morph into specified shapes. … II. Reinforcement Learning Module Reinforcement learning (RL) is a method of learning from interaction between an agent and its environ- … Cited by 9 Related articles View as HTML All 5 versions

[PDF] from itu.dk Modular Reinforcement Learning architectures for artificially intelligent agents in complex game environments CJ Hanna, RJ Hickey, DK Charles… – … Intelligence and Games …, 2010 – ieeexplore.ieee.org … One method of deriving this value is to use the value that an individual reinforcement learning module has built up in its Q table for taking that action in the current state. This allows the action selector to determine which action it expects to produce the maximum global reward. … Cited by 1 Related articles All 5 versions

[PDF] from psu.edu [PDF] Towards a Continuous Reinforcement Learning Module for Navigation in Video Games T Gourdin, O Sigaud – 2009 – Citeseer Abstract. Video games are highly non-stationary environments. Our goal is to build a  navigation module for video games based on Continuous Reinforcement Learning  techniques. A study of the state-of-the-art of these techniques reveals that memory-based … Related articles View as HTML All 7 versions

[PDF] from ifaamas.org Spectrum management of cognitive radio using multi-agent reinforcement learning C Wu, K Chowdhury, M Di Felice… – Proceedings of the 9th …, 2010 – dl.acm.org … How- ever, the concurrent use of the spectrum with a PU 1707 Page 4. (Application Layer) Sensing Spectrum Sharing Reinforcement Learning Module Spectrum Spectrum CR Link Layer Module * * Spectrum Neighbor List * Tx Power CR Physical Layer Module Block … Cited by 9 Related articles All 6 versions

[PDF] from plym.ac.uk Real-world reinforcement learning for autonomous humanoid robot charging in a home environment N Navarro, C Weber, S Wermter – Towards Autonomous Robotic Systems, 2011 – Springer … of sideward movements due to slippage of the Nao. To obtain a more robust solution using this approach, we suggest adding a final module after the reinforcement learning module. The objective of this module will be to check … Cited by 2 Related articles All 8 versions

A Method of Reinforcement Learning Based Automatic Traffic Signal Control W Yaping, Z Zheng – Measuring Technology and Mechatronics …, 2011 – ieeexplore.ieee.org … The input of the control unit is the current state of the intersection, and the output is instruction for each vehicle agent real time. The received current status will be the initial environment state for the reinforcement learning module. … Cited by 2 Related articles All 4 versions

[PDF] from aaaipress.org Cerberus: Applying supervised and reinforcement learning techniques to capture the flag games AS Hefny, AA Hatem, MM Shalaby, AF Atiya – Fourth Artificial Intelligence …, 2008 – aaai.org … ammo. These assumptions do not affect the reinforcement learning module because, as will be shown, it is not concerned with the fineYgrained details of the game state. Also … environments. Reinforcement Learning Module Our … Cited by 5 Related articles All 4 versions

ATFM computational agent based on reinforcement learning aggregating human expert experience AMF Crespo, L Weigang… – … System (FISTS), 2011 …, 2011 – ieeexplore.ieee.org … The device known as Distributed Decision Support System for Air Traffic Flow Management (SISCONFLUX) was built with a computational agent based on Reinforcement Learning (Module MAAD), which uses Q-learning algorithm. … Related articles

Multi Objective Dynamic Job Shop Scheduling using Composite Dispatching Rule and Reinforcement Learning X Chen, XC Hao, HW Lin, T Murata – ??????? C (??·??· …, 2011 – J-STAGE Page 1. © 2011 The Institute of Electrical Engineers of Japan. 1241 ??????? C(??·??·???????) IEEJ Transactions on Electronics, Information and Systems Vol.131 No.6 pp.1241-1249 DOI: 10.1541/ieejeiss.131.1241 … Related articles All 5 versions

Traffic Light Controller Based on Hidden Markov Model and Reinforcement Learning Method M Ali, A Fasih, A Haj Mosa, F Al Machot, K Kyamakya – NDES 2012, 2012 – vde-verlag.de … We use HMM as predictor for estimating traffic volume in an intersection. Then, we feed these data from HMM into the adapted Reinforcement Learning module to choose a best action and control the intersection traffic signal. The objective of this controller is reducing traffic jam. …

Darwinian embodied evolution of the learning ability for survival S Elfwing, E Uchibe, K Doya, HI Christensen – Adaptive Behavior, 2011 – adb.sagepub.com … Nm reinforcement learning modules according to: arg max(wx) (line 7 in Algorithm 2 in Appendix B), where w is the neural network weights matrix (of size Nm  Nx) and x is the current state input vector (of size Nx  1). The selected reinforcement learning module then executes … Cited by 8 Related articles All 8 versions

Dynamic scheduling of maintenance tasks in the petroleum industry: A reinforcement approach N Aissani, B Beldjilali, D Trentesaux – Engineering Applications of Artificial …, 2009 – Elsevier … dynamics of the complex traffic processes within the network. In their model, an on-line reinforcement learning module updates the agents’ knowledge base and inference rules. Their test results showed that the multi-agent system … Cited by 17 Related articles All 3 versions

[HTML] from pnas.org The basal ganglia communicate with the cerebellum AC Bostan, RP Dum, PL Strick – Proceedings of the …, 2010 – National Acad Sciences … Thus, the two subcortical structures may be linked together to form an integrated functional network. One might then ask what new computational operations emerge by interconnecting a reinforcement learning module with a supervised learning module. … Cited by 57 Related articles All 9 versions

[PDF] from osaka-u.ac.jp Efficient Behavior Learning by Utilizing Estimated State Value of Self and Teammates K Shimada, Y Takahashi, M Asada – … 2009: Robot Soccer World Cup XIII, 2010 – Springer … Page 5. Efficient Behavior Learning by Utilizing Estimated State Value 359 An action module of the lower layer has a reinforcement learning module which estimates state values for the action. An agent can discriminate a set S of distinct world states. … Cited by 1 Related articles All 5 versions

Neuro-Fuzzy Actor Critic Reinforcement Learning for determination of optimal timing plans L Chong, M Abbas – Intelligent Transportation Systems (ITSC), …, 2010 – ieeexplore.ieee.org … Three hierarchical layers of controller agents exist: intersection, zone, and regional controller. An online reinforcement learning module updated the knowledge base and rules according to online traffic states by using fuzzy logic with an objective of minimize average delay. … Cited by 1 Related articles All 4 versions

Multi-robot Formation Control Using Reinforcement Learning Method G Zuo, J Han, G Han – Advances in Swarm Intelligence, 2010 – Springer … obstacle environment. Each robot’s control system has a reinforcement learning module to achieve the upper behavior control. Our … ro- bots. This type of sensor information is sent to the reinforcement learning module (Q-learning). The … Related articles All 2 versions

[PDF] from usc.edu Multiple model-based reinforcement learning explains dopamine neuronal activity M Bertin, N Schweighofer, K Doya – Neural Networks, 2007 – Elsevier … Using predictive models, each reinforcement learning module tries to predict the future states. A responsibility signal ? i is assigned to each predictive model i depending on the likelihood of the current observed state and the reliability of the past predictions. … Cited by 10 Related articles All 8 versions

Local Planning of AUV Based on Fuzzy-Q Learning in Strong Sea Flow Field G Yang, R Zhang, D Xu, Z Zhang – … Sciences and Optimization, …, 2009 – ieeexplore.ieee.org … Centroid method. 4. Reinforcement Learning Module 4.1. Q-learning Q-learning (Watkins, 1989) is a form of model-free reinforcement learning. It can also be viewed as a method of asynchronous dynamic programming (DP). It … Related articles All 3 versions

[PDF] from epfl.ch Using reinforcement learning to adapt an imitation task F Guenter, AG Billard – … Robots and Systems, 2007. IROS 2007. …, 2007 – ieeexplore.ieee.org … To avoid that type of problem, we have implemented a reinforcement learning module which allows the robot to learn how to avoid the obstacles or other perturbations. The reinforcement learning acts directly on the modulation of the dynamical system. … Cited by 8 Related articles All 4 versions

[HTML] from nih.gov [HTML] Learning tactile skills through curious exploration L Pape, CM Oddo, M Controzzi, C Cipriani… – Frontiers in …, 2012 – ncbi.nlm.nih.gov … events. Sensory inputs are grouped in an unsupervised manner by the abstractor into separate abstractor states y j . The behavior that leads to each abstractor state y j is learned by an individual reinforcement learning module . …

[PDF] from uni-bremen.de Combining Rule Induction and Reinforcement Learning: An Agent-based Vehicle Routing B Sniezynski, W Wojcik, JD Gehrke… – Machine Learning and …, 2010 – ieeexplore.ieee.org … The AQ rule induction algorithm is used to generate a classifier, which predicts traffic (its discretized value) from date and time data. Prediction given by the classifier, combined with location data, is used as an input to the reinforcement learning module. … Cited by 3 Related articles All 4 versions

[PDF] from ijcai.org Direct code access in self-organizing neural networks for reinforcement learning AH Tan – Proceedings of International Joint Conference on …, 2007 – aaai.org … The clusters are in turns used as the compressed states and actions by a Q-learning module. Ninomiya (2002) couples a supervised ART system with a Temporal Differ- ence reinforcement learning module in a hybrid architecture. … Cited by 16 Related articles All 18 versions

Research of Cooperative Behavior in Multi-Robots System S PIAO, Q ZHONG, Y LIU, Z CAI – Chinese Journal of Electronics, 2011 – 220.194.54.16 … the parametric control and enhancing technique will make the optimal control problem to be transformed into a series of parameter optimization problem, and the decision tree module can accelerate the learning process, so that the reinforcement learning module can choose … Cached All 3 versions

A Role for Posterior Cingulate Cortex in Policy Switching and Cognitive Control JM Pearson, BY Hayden, ML Platt – Neural Basis of Motivational …, 2011 – books.google.com … change detection and policy selection. Sensory feedback from reward outcomes is divided into task-specific variables and passed on to both a reinforcement learning module and a change detector. The learning module computes … Cited by 2 Related articles

In-situ optimal control of nutrient solution for soilless cultivation F Chen, H He, Y Tang – Advanced Computer Control (ICACC), …, 2011 – ieeexplore.ieee.org … Q-learning is one of the most important reinforcement learning algorithms [14]. In reinforcement learning module, an agent acts on environment and receives reinforcement signal which is either punishment or reward induced by environment state transition. … Cited by 1 Related articles

[PDF] from ohio.edu Motivated learning in autonomous systems P Raif, JA Starzyk – … The 2011 International Joint Conference on, 2011 – ieeexplore.ieee.org … The function of the goal creation module is to constitute the hierarchy of new goals and to manage them (and consequently to manage the agent’s behavior according to these goals). Each of the internal goals corresponds to specific reinforcement learning module. … Related articles All 3 versions

[PDF] from umich.edu Instance-Based Online Learning of Deterministic Relational Action Models JZ Xu, JE Laird – Proceedings of the Twenty-Fifth Conference on …, 2010 – aaai.org … Ties are broken by episode recency. Derbinsky & Laird (2009) present evidence that cue-based retrieval is tractable even with one million stored episodes. Soar also has a reinforcement learning module that tunes agent behavior with Q-learning (Nason & Laird 2004). … Cited by 7 Related articles All 6 versions

[PDF] from neu.edu Novel function approximation techniques for large-scale reinforcement learning C Wu – 2010 – iris.lib.neu.edu Page 1. Northeastern University Computer Engineering Dissertations … Related articles All 2 versions

[PDF] from 202.194.20.8 C-NEDAT: A cognitive network engineering design analytic toolset for MANETs L Kant, A McAuley, K Manousakis… – … 2010-MILCOM 2010, 2010 – ieeexplore.ieee.org … multicast trees ? Diverse techniques to enhance OSPF-like / PIM-like protocols ? Multi-Objective (QoS): delay, overhead, capacity, energy, security ? Techniques: greedy heuristic, SA, hierarchical OSPF-area like ? Cognitive Techniques: Reinforcement Learning Module for OLSR … Cited by 5 Related articles All 2 versions

Intelligent computing methods in Air Traffic Flow Management L Weigang, MVP Dib, DP Alves, AMF Crespo – … Research Part C: Emerging …, 2010 – Elsevier … Decision and control approach; 4.4. Reinforcement Learning Module – RLM. 5. Analyses of two systems; 5.1. Discussion about the ATFMGC; 5.2. Discussion about the SISCONFLUX. 6. Conclusions and further research; Acknowledgements; References. 1. Introduction. … Cited by 5 Related articles All 4 versions

A Fuzzy-Q Method to Improve the Adaptability of AUV in Variable Ocean Current Environment G Yang, R Zhang, D Xu – … Systems, 2009. GCIS’09. WRI Global …, 2009 – ieeexplore.ieee.org … Centroid method. IV. REINFORCEMENT LEARNING MODULE Q-learning (Watkins, 1989) is a form of model-free reinforcement learning. It can also be viewed as a method of asynchronous dynamic programming (DP). It provides … Cited by 1 Related articles All 3 versions

[PDF] from uni-bonn.de Improving imitated grasping motions through interactive expected deviation learning K Gra¨ve, J Stu¨ckler, S Behnke – … Robots (Humanoids), 2010 …, 2010 – ieeexplore.ieee.org … single trial. In more recent work, Billard and Guenter [13] extended their imitation learning framework by a reinforcement learning module, in order to be able to handle unexpected changes in the envi- ronment. However, both … Cited by 2 Related articles All 7 versions

[HTML] from frontiersin.org Frontiers L Pape, CM Oddo, M Controzzi, C Cipriani… – Frontiers in …, 2012 – frontiersin.org Frontiers | Learning tactile skills through curious exploration | Frontiers in Neurorobotics publishes articles on the most outstanding discoveries across the research spectrum of Frontiers | Learning tactile skills through curious exploration | Neurorobotics. Cached

User to user QoE routing system H Tran, A Mellouk – Wired/Wireless Internet Communications, 2011 – Springer … The idea of applying RL to routing in networks was firstly introduced by Boyan and Littman [4]. Authors described the Q routing algorithm for packet routing. Reinforcement learning module is integrated into each node in net- work. … Related articles All 2 versions

A Mixed Reality Based Teleoperation Interface for Mobile Robot X Wang, J Zhu – Mixed Reality and Human-Robot Interaction, 2011 – Springer … (2005) have developed reinforcement learning as a computational approach of automating goal-directed learning and decision-making. The reinforcement learning module could help semi-autonomous robot to learn which action it should take in response to a state or rewards. … Related articles All 2 versions

User QoE-based adaptive routing system for future Internet CDN HA Tran, A Mellouk, S Hoceini… – Computing, …, 2012 – ieeexplore.ieee.org … The idea of applying reinforcement learning to routing in networks was firstly introduced by [4]. Authors described the Q-routing algorithm for packet routing. Reinforcement learning module is embedded into each node of a switching network. …

Use of machine learning for continuous improvement of the real time heterarchical manufacturing control system performances N Aissani, B Beldjilali, D Trentesaux – International Journal of Industrial …, 2008 – Inderscience … (2003) used reinforcement learning for real-time coordinated signal control in an urban traffic network, they used a multiagent approach and to handle the changing dynamics of the complex traffic processes within the network, an online reinforcement learning module is used to … Cited by 9 Related articles All 5 versions

[PDF] from ntnu.no [CITATION] Learning Trust in Dynamic Multiagent Environments using HMMs MEG Moe, M Tavakolifard, SJ Knapskog – Proceedings of the 13th Nordic Workshop …, 2008 Cited by 5 Related articles All 6 versions

[PDF] from uminho.pt An intelligent alarm management system for large-scale telecommunication companies R Costa, N Cachulo, P Cortez – Progress in Artificial Intelligence, 2009 – Springer … The evaluation metrics include sup- port, lift, support and information on reception order. These statistics are up- dated every time new information is collected. • Reinforcement Learning Module – refines the rules database based on feedback obtained from the Portal Server. … Cited by 5 Related articles All 5 versions

Real-Time state-dependent routing based on user perception HA Tran, A Mellouk – Communications and Information …, 2011 – ieeexplore.ieee.org … The idea of applying reinforcement learning to routing in networks was firstly introduced by [8]. Authors described the Q-routing algorithm for packet routing. Reinforcement learning module is embedded into each node of a switching network. … Cited by 3 Related articles

Incremental acquisition of behaviors and signs based on a reinforcement learning schemata model and a spike timing-dependent plasticity network T Taniguchi, T Sawaragi – Advanced Robotics, 2007 – ingentaconnect.com … We found that spike timing-dependent plasticity (STDP) can be used to learn the ordinal relationships of incoming SD and to select the next required reinforcement learning module (see Section 3). 1.4. Emergence of behaviors and signs … Cited by 9 Related articles All 3 versions

Context-aware dynamic service composition in ubiquitous environment K Tari, Y Amirat, A Chibani, A Yachir… – … (ICC), 2010 IEEE …, 2010 – ieeexplore.ieee.org … C.2.1. QoS Monitoring The proposed architecture uses Q-learning as reinforcement-learning technique because it provides good experimental results in terms of learning speed [9]. The reinforcement-learning module can be quite reactive to decide which next action to take. … Cited by 6 Related articles

[PDF] from ucsd.edu A robotic model of the development of gaze following H Kim, H Jasso, G Deák… – Development and Learning …, 2008 – ieeexplore.ieee.org … h, shown in Fig. 4. This state vector is fed to the reinforcement learning module and the module computes which action a to take to maximize its reward, which is just the saliency of the object itself. Once an action is complete … Cited by 7 Related articles All 5 versions

[PDF] from diva-portal.org [PDF] Paper E MEG Moe, M Tavakolifard… – Security, Privacy and …, 2009 – hkr.diva-portal.org … reinforcement learning. By this combination, the reliability of the hidden Markov model will be improved since its parameters are re-estimated after training of the model with the reinforcement learning module. 1. Introduction … Related articles View as HTML All 4 versions

[PDF] from aucmcs.ro [PDF] Adaptive Web Recommendation Systems M Preda, AM Mirea, C Teodorescu-Mihai… – Annals of the University …, 2009 – aucmcs.ro … 25 Page 2. 26 M. PREDA, A.-M. MIREA, C. TEODORESCU-MIHAI, AND DL PREDA Web site Reinforcement Learning Module Web usage, feedback (rewards) Recommendations Ontological structure of the Web site content Simil arity measure … Cited by 2 Related articles View as HTML All 5 versions

Evolutionary Robotics: From Simulation-Based Behavior Learning to Direct Teaching in Real Environments S Yamada, D Katagami – Design and Control of Intelligent Robotic …, 2009 – Springer Page 1. 4 Evolutionary Robotics: From Simulation-Based Behavior Learning to Direct Teaching in Real Environments Seiji Yamada1 and Daisuke Katagami2 1 National Institute of Informatics seiji@nii.ac.jp 2 Tokyo Institute of Technology katagami@ntt.dis.titech.ac.jp Abstract. … Related articles All 2 versions

[PDF] from thinkmind.org User to User adaptive routing based on QoE HA Tran, A Mellouk, S Hoceini – Programming and Systems ( …, 2011 – ieeexplore.ieee.org … The idea of applying reinforcement learning to routing in networks was firstly introduced by [3]. Authors described the Q-routing algorithm for packet routing. Reinforcement learning module is embedded into each node of a switching network. … Related articles All 2 versions

[PDF] from nju.edu.cn [PDF] XCSc: A novel approach to clustering with extended classifier system LD SHI, YH SHI, Y GAO, LIN SHANG… – International Journal of …, 2011 – cs.nju.edu.cn … 7. 2. Introduction to XCS Generally, a LCS consists of two key components: (i) a reinforcement learning module for receiving the feedback of rewards from the environment and updating the rule parameters; and (ii) a genetic algorithm module for evolving the rule population. … Cited by 8 Related articles View as HTML All 3 versions

Real-world production scheduling for the food industry: An integrated approach T Wauters, K Verbeeck, P Verstraete… – … Applications of Artificial …, 2011 – Elsevier Cited by 2 Related articles All 3 versions

[PDF] from ijac.net Coordination control of greenhouse environmental factors F Chen, YN Tang, MY Shen – International Journal of Automation and …, 2011 – Springer … In reinforcement learning module, an agent acts on environment and receives reinforcement signal which is either punishment or reward induced by environment state transition. The learning task for an agent is to search for an … Related articles All 6 versions

[PDF] from osaka-u.ac.jp Efficient behavior learning based on state value estimation of self and others Y Takahashi, K Noma, M Asada – Advanced Robotics, 2008 – ingentaconnect.com … 2.2. Action Module An action module of the lower layer has a reinforcement learning module which estimates state values for the action. Figure 2 shows a basic model of reinforcement learning. An agent can discrimi- nate a set S of distinct world states. … Cited by 3 Related articles All 3 versions

Towards constraint optimal control of greenhouse climate F Chen, Y Tang – Life System Modeling and Intelligent Computing, 2010 – Springer … presented by Brooks [21]. In reinforcement learning module, an agent acts on environment and receives reinforcement signal which is either punish- ment or reward induced by environment state transition. The learning task … Cited by 2 Related articles All 3 versions

An attempt to obtain scheduling rules of network-based support system for Decentralized Scheduling of Distributed Production Systems H Yamaha, S Matsumoto… – Industrial Informatics, 2008. …, 2008 – ieeexplore.ieee.org … 7). A combination of each of the three actions and a state forms a “rule”. The module of 509 Page 5. Module of Reinforcement Learning State i Action 1 Action 2 Action 3 Rule i-1 Rule i-2 Rule i-3 wi-1 wi-2 wi-3 weight Production System Module of Reinforcement Learning … Cited by 2 Related articles

An architecture for learning agents B Sniezynski – Computational Science–ICCS 2008, 2008 – Springer … Additional research is necessary to provide guidance in this aspect. Two types of learning modules were developed and tested so far: reinforcement learning module and inductive rule learning (see section 4). Although, other learning methods can be also used. … Cited by 2 Related articles BL Direct All 2 versions

Intelligent Searching Methods for Open Source Information in Emergency Logistics’ Decision System H Song – e-Business and Information System Security (EBISS), …, 2010 – ieeexplore.ieee.org … descriptive mode of records. Thereafter, the needed records were taken out and saved to the database of information memory for user’s enquiry. • The reinforcement learning module for searching knowledge. Accepting the success or … Related articles

[PDF] from tamu.edu [PDF] Reinforcement Learning for Active Length Control of Shape Memory Alloys K Kirkpatrick, J Valasek – & Proceedings ??· ?????| ????? …, 2008 – vscl.tamu.edu … t t v v H ad ? ? + ? (13) Note that the purpose of the Reinforcement Learning module is not to compute the exact Q(s,a), but to learn the optimal policy with the optimal action-value function. In the process of policy iteration, an intermediate policy can … Cited by 1 Related articles View as HTML All 5 versions

[PDF] from osaka-u.ac.jp Cooperative/competitive behavior acquisition based on state value estimation of others K Noma, Y Takahashi, M Asada – … 2007: Robot Soccer World Cup XI, 2008 – Springer … 103 Fig. 1. A multi-module learning system 2.2 Action Module An action module of the lower layer has a reinforcement learning module which estimates state values for the action. Fig. 2. Agent-environment interaction Fig. 3. Sketch of a state value function … Cited by 3 Related articles All 4 versions

Profile-Based Adaptive DiffServ Policing with Learning Techniques L Cruvinel, T Vazao – Computer Communications and …, 2011 – ieeexplore.ieee.org … Algorithm 1 Reinforcement Learning Module Reinforcement Learning module 1 Read or Initialize Knowledge Base 2 Main cycle: 3 Wait for Notification or Timeout 4 if Notification 5 Get state(Notification,Notification interval, Non Priority traffic sent) 6 Choose action(state) 7 Apply … Related articles

[PDF] from tamu.edu [PDF] Reinforcement Learning for Characterizing Hysteresis Behavior of Shape Memory Alloys K Kirkpatrick, J Valasek – & Proceedings ??· ?????| ?? …, 2007 – jungfrau.tamu.edu … for updating ?v becomes t t v v H ad ? ? + ? (13) Note that the purpose of the Reinforcement Learning module is not to compute the exact Q (s,a), but to learn the optimal policy with the optimal action-value function. In the process of policy iteration, an intermediate policy can … Cited by 6 Related articles View as HTML All 5 versions

[PDF] from ul.pt Adaptability Support in a Learning Management System LCPL Santos – 2009 – docs.di.fc.ul.pt … 39 – 4.4 Reinforcement Learning Module ….. … Implementation of a platform independent tutoring module (reinforcement learning module) that provides an HTTP interface in order to make it portable to other e-learning … Cited by 1 Related articles All 2 versions

[PDF] from ntu.edu.sg Intelligence through interaction: Towards a unified theory for learning AH Tan, G Carpenter, S Grossberg – Advances in Neural Networks–ISNN …, 2007 – Springer … manner. Compared with other ART-based reinforcement learn- ing systems, FALCON presents a truly integrated solution in the sense that there is no implementation of a separate reinforcement learning module or Q-value ta- ble. … Cited by 25 Related articles BL Direct All 4 versions

[PDF] from nju.edu.cn Clustering with xcs on complex structure dataset L Shi, Y Gao, L Wu, L Shang – AI 2008: Advances in Artificial Intelligence, 2008 – Springer … LCS mainly consists of two important components: (1) reinforcement learning module, for receiving rewards feed back from the environment and updating the parameters of classifier rules; (2) genetic algorithm module, for generating new classifier rules and evolving the rule … Cited by 4 Related articles All 4 versions

[PDF] from psu.edu Personalized web content provider recommendation through mining individual users’ QoS S Xu, H Jiang, F Lau – Proceedings of the 11th International Conference …, 2009 – dl.acm.org … For ex- ample, Boyan and Littman [1] proposed a Q-routing algorithm by embedding a reinforcement learning module into each node in the routing network for minimizing the delivery time of data packages over the switching network. … Related articles All 5 versions

Unified service conflict control in IMS L Hua, Z Xue, Y Fangchun – … , 2009. ConTEL 2009. 10th …, 2009 – ieeexplore.ieee.org … in AIS, is used to build up service interaction models out of the incoming signaling messages; 2)Conflict Recognition Module ,which simulates Antigen Recognition in AIS, serves to manage various conflicts uniformly; 3) Reinforcement Learning Module implements Immune … Related articles

[PDF] from organic-computing.org [PDF] Adaptive object acquisition G Peters, T Leopold, CP Alberts, M Briese… – … et al.[4], 2008 – organic-computing.org … Universität Dortmund, Informatik VII, D-44227 Dortmund, Germany Abstract We propose an active vision system for object acquisition. The core of our approach is a reinforcement learning module which learns a strategy to scan an object. … Cited by 1 Related articles View as HTML All 11 versions

LCSE: learning classifier system ensemble for incremental medical instances Y Gao, J Huang, H Rong, D Gu – Learning Classifier Systems, 2007 – Springer … The classical LCS includes two major modules, a genetic algorithm module used to facilitate rule discovery, and a reinforcement learning module used to adjust the strength of the corresponding rules after the learning module receives the rewards from the environment. … Cited by 5 Related articles BL Direct All 2 versions

Biologically inspired framework for learning and abstract representation of attention control H Fatemi Shariatpanahi, M Nili Ahmadabadi – Attention in Cognitive …, 2007 – Springer … 5.1 Learning Control of Attention In this phase, a reinforcement learning module in the agent’s brain invents strategies for concurrent deployment of motor actions and attention shifts, in order to maximize its average receiving reward from the environment. … Cited by 11 Related articles BL Direct All 3 versions

[BOOK] A comparative study of Roth-Erev and modified Roth-Erev reinforcement learning algorithms for uniform-price double auctions M Pentapalli – 2008 – books.google.com Page 1. A comparative study of Roth-Erev and modi?ed Roth-Erev reinforcement learning algorithms for uniform-price double auctions by Mridul Pentapalli A thesis submitted to the graduate faculty in partial ful?llment of the … Cited by 5 Related articles Library Search All 3 versions

[PDF] from psu.edu [PDF] Interactive learning of top-down attention control and motor actions A Borji, MN Ahmadabadi, BN Araabi – … on From motor to interaction learning …, 2008 – Citeseer … interacting modules. A reinforcement learning module learns a policy on a set of regions in the room for reaching the target object, using as objective function which is the expected value of the sum of discounted rewards. By … Cited by 2 Related articles View as HTML All 4 versions

A bio-inspired architecture of an active visual search model V Cutsuridis – Artificial Neural Networks-ICANN 2008, 2008 – Springer … See text for the corre- sponding to the model’s modular functionality of each brain area. Page 3. 250 V. Cutsuridis decision-making module, a reinforcement learning module, a motor plan module and a motor execution module. … Reinforcement learning module. … Cited by 2 Related articles All 6 versions

[PDF] from socionet.org Dynamic LMP response under alternative price-cap and price-sensitive demand scenarios H Li, J Sun, L Tesfatsion – … and Delivery of Electrical Energy in …, 2008 – ieeexplore.ieee.org … solves its DC-OPF problems by invoking DCOPFJ. Generator learning is implemented in the AMES test bed by a reinforcement learning module, JReLM, developed by Gieseler [10]. JReLM can implement a variety of different … Cited by 11 Related articles All 15 versions

Learning classifier system ensemble and compact rule set Y Gao, JZ Huang, L Wu – Connection Science, 2007 – Taylor & Francis … Each learning classifier system in the first level consists of two major modules, a genetic algorithm module for facilitating rule-discovery and a reinforcement learning module for adjusting the strength of the corresponding rules when rewards are received from the environment. … Cited by 8 Related articles BL Direct All 5 versions

[PDF] from psu.edu An agent-based computational laboratory for wholesale power market design J Sun, L Tesfatsion – Power Engineering Society General …, 2007 – ieeexplore.ieee.org … Trader learning is implemented in the AMES framework by a reinforcement learning module, JReLM, developed by Gieseler [7]. JReLM can implement a variety of different reinforcement learning methods, permitting flexible repre- Page 3. Fig. … Cited by 13 Related articles All 10 versions

[CITATION] On-line scheduling of Automatics and flexible Manufacturing System using SARSA technique N Aissani, B Beldjilali, H Arioui, R Merzouki… – Aip Conference …, 2008 Related articles All 2 versions

Vision-based Navigation and Reinforcement Learning Path Finding for Social Robots X Pérez Sala – 2011 – upcommons.upc.edu … Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial::Aprenentatge automàtic Àrees temàtiques de la UPC::Informàtica::Robòtica Reinforcement learning Robotics Artificial Vision module Behavior control module Reinforcement Learning module Robot Navigation … Cached All 2 versions

[PDF] from wits.ac.za [PDF] HONOURS RESEARCH PROPOSAL B SHONGWE – 2011 – cs.wits.ac.za … Both the transcribed audio file and the original music audio file will be taken into the reinforcement learning module. The reinforcement learning module will compare the similarity of the original music audio and the transcribed audio piece. … Related articles View as HTML

[PDF] from tudelft.nl [PDF] Avoiding failure states during Reinforcement Learning M Van Diepen, IE Schuitema, W Caarls – 2011 – repository.tudelft.nl … determined. When the supervisor is also a Reinforcement Learning module, the learning time of the whole system is likely to increase and the advantage during learning may vanish. Hence, this method was not implemented. … Related articles View as HTML All 2 versions

[PDF] from tamu.edu [PDF] Space-Based Antenna Morphing using Reinforcement Learning HA Feldman – & Proceedings ??· ?????| ?????| ???? …, 2007 – tiims.tamu.edu … which employs Q-Learning and an e-greedy method to ensure sufficient information is acquired by the Reinforcement Learning module. … Instead of relying on user input to change the shape of the antenna model, the Reinforcement Learning module … Cited by 1 Related articles View as HTML All 5 versions

[PDF] from skemman.is [PDF] A Distributed Dialogue Architecture with Learning GR Jónsdóttir – 2008 – skemman.is Page 1. M.Sc. Thesis Reykjavík University – School of Computer Science Guðný Ragna Jónsdóttir Master of Science November 2008 A Distributed Dialogue Architecture With Learning Page 2. Page 3. Thesis Committee: Dr. Kristinn … Cited by 1 Related articles View as HTML All 2 versions

[PDF] from ucsd.edu iMime: an interactive character animation system for use in dementia care A Wiratanaya, MJ Lyons, NJ Butko, S Abe – Proceedings of the 12th …, 2007 – dl.acm.org … functional prototype. Preliminary tests with naïve users showed that the system is stable and works as intended. The reinforcement learning module, however, adapts rather slowly in response to user behaviour. Future work … Cited by 2 Related articles All 6 versions

[PDF] from diva-portal.org [PDF] Security, Privacy and Trust in Dynamic Networks MEG Moe – 2009 – umu.diva-portal.org … The former and latter are constructed from hidden Markov modeling and reinforcement learning (RL), as illustrated in Figure 4. The model parameters of the HMM are re-estimated after having learnt about its environment from the reinforcement learning module. … View as HTML All 6 versions

Mining knowledge from data using anticipatory classifier system O Unold, K Tuszynski – Knowledge-Based Systems, 2008 – Elsevier … Mutation is a process that generalizes condition part of classifiers. If the size of the action set is too big, some classifiers are deleted in order to make space for newly generated ones. ACS also includes reinforcement learning module which uses the Q-learning idea. … Cited by 7 Related articles All 2 versions

[PDF] from nus.edu.sg Multi-Robot Concurrent Learning in Museum Problem LIU Zheng, MH Ang, WKG Seah – Distributed Autonomous Robotic …, 2007 – Springer … Instead, the reinforcement learning module adjusts the weight inside the network through the interaction with the environment. … It should be noted that the input/output spaces of the reinforcement learning module are still discrete and finite. … Related articles All 4 versions

Neural networks in automotive applications D Prokhorov – Computational Intelligence in Automotive Applications, 2008 – Springer … functional if necessary), or without a model. Model-free adaptive control is implemented sometimes with the help of a reinforcement learning module called critic [54, 55]. Applications of reinforcement learning and approximate … Cited by 4 Related articles All 2 versions

[PDF] from psu.edu Dynamic learning of action patterns for object acquisition G Peters, T Leopold – … Journal of Intelligent Systems Technologies and …, 2007 – Inderscience … Thus, a state of the reinforcement learning module consists of six components: x-position on the hemisphere (20 possible values), y-position (five possible values), and unfamiliarity of the areas in the four directions (five possible values each), resulting in a total of 2000 states. … Cited by 1 Related articles BL Direct All 15 versions

Special Issue on Evolutionary Learning and Optimisation X Li, W Luo, X Yao – 2007 – Taylor & Francis … The first level contains a set of LCSs which are trained with different data sets re-sampled from a training data set, where a genetic algorithm module is used to facilitate rule discovery and a reinforcement learning module is used to adjust the strength of the rules. … BL Direct All 5 versions

[PDF] from ucr.edu [PDF] Learning to Perceive for Autonomous Navigation in Outdoor Environments J Peng, B Bhanu – 2007 – vislab.ee.ucr.edu … Reinforcement learning module then uses this confi- dence value as feedback to induce a mapping from images to segmentation strategies within each local region created from partitioning of the feature space. The goal is, therefore, to maximize the matching con- 97 Page 4. … Related articles All 8 versions

[PDF] from drdc.gc.ca [PDF] Shape-shifting Tracked Robotic Vehicle for complex terrain navigation I Vincent, M Trentini – Defence R&D Canada, Technical …, 2007 – pubs.drdc.gc.ca … Figure 6 presents the main idea of the learning architecture. The first control layer inputs the terrain map into a reinforce- ment learning algorithm for the axle angular position computation. Each behaviour has a first layer reinforcement learning module trained individually. … Cited by 2 Related articles

[PDF] from 129.186.33.9 Dynamic testing of wholesale power market designs: An open-source agent-based framework J Sun, L Tesfatsion – Computational Economics, 2007 – Springer … As detailed in Sect. 3.6 below, trader learning is implemented in the AMES framework by a reinforcement learning module,JReLM, developed by Gieseler (2005). JReLM can implement a variety of different reinforcement learning methods, permitting … Cited by 84 Related articles BL Direct All 24 versions

[PDF] from iastate.edu [PDF] Learning Algorithms: Illustrative Examples L Tesfatsion – link at Syllabus Section III. B, directly accessible at, 2007 – econ.iastate.edu Page 1. 1 Learning Algorithms: Illustrative Examples Notes for Economics 308 (Agent-Based Computational Economics) Leigh Tesfatsion Professor of Economics and Mathematics Iowa State University Ames, IA 50011-1070 http://www.econ.iastate.edu/tesfatsi/ … Cited by 2 Related articles View as HTML All 6 versions

[PDF] from buffalo.edu [PDF] Ph. D. Dissertation Proposal MGLAIR: A Multimodal Cognitive Agent Architecture J Bona – 2010 – cse.buffalo.edu … itself. Soar [Laird et al., 1987] has its roots in traditional, symbolic production systems but has been extended to include modules for semantic and episodic learning and memory, a reinforcement learning module, and more. Recent … Related articles View as HTML All 2 versions

[PDF] from ualberta.ca [PDF] Module-Based Reinforcement Learning: Experiments with a Real Robot CS Ari – 2007 – ualberta.ca … environments. Keywords: reinforcement learning, module-based RL, robot learning, problem decomposition, Markovian Decision Problems, feature space, subgoals, local control, switching control 1. Introduction Reinforcement … Related articles View as HTML All 2 versions

[PDF] from psu.edu Agent-based simulation of electricity markets: a survey of tools Z Zhou, WKV Chan, JH Chow – Artificial Intelligence Review, 2007 – Springer Page 1. Artif Intell Rev (2007) 28:305–342 DOI 10.1007/s10462-009-9105-x Agent-based simulation of electricity markets: a survey of tools Zhi Zhou · Wai Kin (Victor) Chan · Joe H. Chow Published online: 12 July 2009 © Springer Science+Business Media BV 2009 … Cited by 26 Related articles All 11 versions

Dopamine and reward prediction error: an axiomatic approach to neuroeconomics A Caplin, M Dean – 2007 – papers.ssrn.com … If we can determine that dopamine does form part of a reinforcement learning module in the brain, understanding how it works can help us to answer several important questions. When is the reinforcement learning system active? … Cited by 5 Related articles

[PDF] from psu.edu [PDF] A COGNITIVE ROBOTIC SYSTEM BASED ON THE SOAR SD Hanford – 2011 – etda.libraries.psu.edu Page 1. The Pennsylvania State University The Graduate School A COGNITIVE ROBOTIC SYSTEM BASED ON THE SOAR COGNITIVE ARCHITECTURE FOR MOBILE ROBOT NAVIGATION, SEARCH, AND MAPPING MISSIONS A Dissertation in Aerospace Engineering … View as HTML

[PDF] from lau.edu.lb SDDSR.(c2005) W Kdouh – 2011 – ecommons.lau.edu.lb Page 1. ~v.j. i’/1’3 Q “I 1 Ci 7 SDDSR: Sequence Driven Dynamic Source Routing for Ad hoe Mobile Networks by Wael Kdouh Submitted in Partial Ful?llment of the Requirements for the Degree of Masters of Science in Computer … Related articles

[PDF] from unige.it [PDF] Modellistica e simulazione ad agente dei liberi mercati dell’energia elettrica LM Benvenuto – 2009 – elettronica.unige.it Page 1. Università degli Studi di Genova Facoltà di Ingegneria Tesi di Laurea in Ingegneria Elettronica Modellistica e simulazione ad agente dei liberi mercati dell’energia elettrica Candidato: Luigi Mauro Benvenuto Relatore: Chiar.mo Prof. Ing. Silvano Cincotti … Cited by 1 Related articles View as HTML All 3 versions