2,821 Hits in 5.9 sec

Use Knowledge to Learn Faster: Topology Recognition with Advice [chapter]

Emanuele Guido Fusco, Andrzej Pelc, Rossella Petreschi
2013 Lecture Notes in Computer Science  
We investigate tradeoffs between the time in which topology recognition is accomplished and the minimum size of advice that has to be given to nodes.  ...  Our results show how sensitive is the minimum size of advice to the time allowed for topology recognition: allowing just one round more, from D to D + 1, decreases exponentially the advice needed to accomplish  ...  In particular, it would be interesting to find the minimum time in which topology recognition can be accomplished using advice of constant size, or even of size exactly 1.  ... 
doi:10.1007/978-3-642-41527-2_3 fatcat:xuyaizjd6nb57nsmzj7bvhpyaq

Evolving neural networks

Risto Miikkulainen
2007 Proceedings of the 2007 GECCO conference companion on Genetic and evolutionary computation - GECCO '07  
is useful, it stays • Initial behaviors, on-line advice • Injecting human knowledge as rules • DEMO Lessons from NERO • NEAT is a strong method for real-time adaptation -Complex team behaviors can be  ...  to train • Improves evolution of a value function -Faster than NEAT alone, better than Q-learning • Utilize both evolution and on-line learning 30 No Targets: Unsupervised Learning • Hebbian  ... 
doi:10.1145/1274000.1274119 dblp:conf/gecco/Miikkulainen07 fatcat:4zyvbu3dxjb4hjqqicvwwzvjna

How Machine (Deep) Learning Helps Us Understand Human Learning: the Value of Big Ideas [article]

Marc Maliar
2019 arXiv   pre-print
I use simulation of two multilayer neural networks to gain intuition into the determinants of human learning.  ...  The first network, the teacher, is trained to achieve a high accuracy in handwritten digit recognition. The second network, the student, learns to reproduce the output of the first network.  ...  In that case, learning from the data is faster than learning from the teacher, which leads us to the following conclusion.  ... 
arXiv:1903.03408v2 fatcat:5a5477f3erb3bpu7v7xjkef4qa

A Gesture Learning Interface for Simulated Robot Path Shaping With a Human Teacher

Paul M. Yanik, Joe Manganelli, Jessica Merino, Anthony L. Threatt, Johnell O. Brooks, Keith Evan Green, Ian D. Walker
2014 IEEE Transactions on Human-Machine Systems  
Experimental results using an automated reward are presented that compare learning results involving single nodes versus results involving the influence of node neighborhoods.  ...  Typically, gesture recognition takes the form of template matching in which the human participant is expected to emulate a choreographed motion as prescribed by the researchers.  ...  , and learning goal configurations with no prior knowledge of the user's preferences.  ... 
doi:10.1109/tsmc.2013.2291714 fatcat:gnrr22r24zclzhkgxpx6vsw7li

Adaptive integrated image segmentation and object recognition

B. Bhanu, Jing Peng
2000 IEEE Transactions on Systems Man and Cybernetics Part C (Applications and Reviews)  
Therefore, it is used in conjunction with model matching where the matching confidence is used as a reinforcement signal to provide optimal segmentation evaluation in a closed-loop object recognition system  ...  Segmentation parameters are represented by a team of generalized stochastic learning automata and learned using connectionist reinforcement learning techniques.  ...  ACKNOWLEDGMENT The authors would like to thank X. Bao for providing us with the experimental data.  ... 
doi:10.1109/5326.897070 fatcat:ckwoby7dzfgwzp7xhy4j6zd4je

Behavior construction and refinement from high-level specifications

Andrew J. Martignoni III, William D. Smart, Douglas W. Gage
2004 Mobile Robots XVII  
We will also show how this can be done automatically, using reinforcement learning, and point out the problems that must be overcome for this approach to work. * The Const component takes one parameter  ...  Often robots are loosely supervised by humans who are not intimately familiar with the inner workings of the robot.  ...  Even these approaches tend to present a fixed solution once the learning phase is complete. Others tried to use advice to help the training speed.  ... 
doi:10.1117/12.568865 dblp:conf/mr/MartignoniS02 fatcat:ldtmee2i65agfpisrsq2mkrjzi

Norm emergence in multiagent systems: a viewpoint paper

Andreasa Morris-Martin, Marina De Vos, Julian Padget
2019 Autonomous Agents and Multi-Agent Systems  
In consequence, we seek to analyse and categorize the approaches proposed in the literature for facilitating norm emergence. This paper makes three contributions to the study of norm emergence.  ...  Firstly, we present the different perspectives of norms and their impact on the norm emergence process, with the aim of comparing their similarities and differences in implementing the norm life cycle.  ...  , and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.  ... 
doi:10.1007/s10458-019-09422-0 fatcat:aqqcpud7mzcutkp4xotzygz46e

A contemporary study into the application of neural network techniques employed to automate CAD/CAM integration for die manufacture

Lian Ding, Jason Matthews
2009 Computers & industrial engineering  
Artificial neural networks have a proven track record in pattern recognition and there ability to learn seems to offer an approach to aid both feature recognition and process planning tasks.  ...  Core to this process is feature recognition.  ...  ., 2001) , used sets of scanned points to recognise features associated with prismatic parts. Wong and Lam (2000) used the topology to recognize orthogonal and non-orthogonal machined features.  ... 
doi:10.1016/j.cie.2009.01.006 fatcat:qjktoeanqrgwblgj33ruapdfqu

Robust convention emergence in social networks through self-reinforcing structures dissolution

Daniel Villatoro, Jordi Sabater-Mir, Sandip Sen
2013 ACM Transactions on Autonomous and Adaptive Systems  
Our main goal is to provide agents with tools that allow them to leverage their social network of interactions while effectively addressing coordination and learning problems, paying special attention  ...  This result leads us to perform an exhaustive analysis of irregular networks discovering what we have defined as Self-Reinforcing Structures (SRS).  ...  With these formalizations and identification of what the frontiers are in each of the topologies, agents can be equipped with this knowledge and the tools necessary to recognize when they are located in  ... 
doi:10.1145/2451248.2451250 fatcat:egm7qyxx4rftbjwuf2j36iuome

Enhancing network modularity to mitigate catastrophic forgetting

Lu Chen, Masayuki Murata
2020 Applied Network Science  
AbstractCatastrophic forgetting occurs when learning algorithms change connections used to encode previously acquired skills to learn a new skill.  ...  From the results, the obtained neural network has a highly modular structure and can learn an unlearned goal faster than without this method.  ...  Thanks to Nadav Kashtan for providing his source code.  ... 
doi:10.1007/s41109-020-00332-9 fatcat:3wqc7beworfg5lmafo6aqjowu4

Learning to Generalize from Single Examples in the Dynamic Link Architecture

Wolfgang Konen, Christoph von der Malsburg
1993 Neural Computation  
We demonstrate that style with a system that learns to classify mirror symmetric pixel patterns from single examples.  ...  The power of this mechanism is due to the very general a priori principle of conservation of topological structure.  ...  One widespread piece of advice is to use input representations that are already adapted to the problem at hand.  ... 
doi:10.1162/neco.1993.5.5.719 fatcat:27xro6gkknhepjzwkd6hr55rau


2020 Journal of Xidian University  
Expert System, development of an Expert System, tools for an Expert System development and advantages, limitations of an Expert System and Decision Support System to Decision Makers.  ...  This Paper deals with the basic concepts of Expert System, literature review, purpose and need of an Expert System, characteristics and categories of an Expert System, architecture and components of an  ...  MLS are frequently used for learning neural network, knowledge bases, using inductive rule based algorithms and evolution-based genetic algorithms, etc.  ... 
doi:10.37896/jxu14.4/131 fatcat:6pvygoynmzgmrls724j7xbtxfq

Creating Advice-Taking Reinforcement Learners [chapter]

Richard Maclin, Jude W. Shavlik
1998 Learning to Learn  
Subsequent reinforcement learning further integrates and refines the advice.  ...  Finally, we present experimental results that indicate our method is more powerful than a naive technique for making use of advice.  ...  We also wish to thank Carolyn Allex, Mark Craven, Diana Gordon, Leslie Pack Kaelbling, Sebastian Thrun, and the two anonymous reviewers for their helpful comments on drafts of this paper.  ... 
doi:10.1007/978-1-4615-5529-2_13 fatcat:u7gr7clcajgszlucwe52c7jq24

Creating advice-taking reinforcement learners

Richard Maclin, Jude W. Shavlik
1996 Machine Learning  
Subsequent reinforcement learning further integrates and refines the advice.  ...  Finally, we present experimental results that indicate our method is more powerful than a naive technique for making use of advice.  ...  We also wish to thank Carolyn Allex, Mark Craven, Diana Gordon, Leslie Pack Kaelbling, Sebastian Thrun, and the two anonymous reviewers for their helpful comments on drafts of this paper.  ... 
doi:10.1007/bf00114730 fatcat:gxsnd5sw75bgxaswkmjghjb77q

The time-sliced paradigm—a connectionist method for continuous speech recognition

Ingrid Kirschning, Hideto Tomabechi, Masafumi Koyama, Jun-Ichi Aoe
1996 Information Sciences  
This is a method for the analysis of the speech signal with the aim to achieve the recognition of connected speech with less preprocessing of the input signal than other existing neural networks.  ...  This network uses a parallel-modular version of Fahlman's Recurrent Cascade-Correlation Learning Architecture (RCC).  ...  Acknowledgments Thanks to Scott Fahlman, Alex Waibel, Gerardo Ayala and Julio Tanomaru for their advice.  ... 
doi:10.1016/0020-0255(96)00083-7 fatcat:tvdeb6djtbdkbdlp3pho67isve
« Previous Showing results 1 — 15 out of 2,821 results