Alexey Vasilyev Transport, Alexey Vasilyev · 2002
Impetuous development of artificial neural networks makes it possible to transfer many ideas from this area into adjacent areas. This work investigates an opportunity of mapping learning classifier systems (LCS) into artificial neural networks (ANN). Possible learning types for hybrid connectionist classifier system (CLCS) for multi-step problems are derived. Transformation's opportunity of the existing LCS into CLCS is shown. Heuristic LCS and CLCS comparisons in the dynamic environment show superiority of CLCS over LCS.
Cited by 0
Alexey Vasilyev · 2002
Abstract: Impetuous development of artificial neural networks makes it possible to transfer many ideas from this area into adjacent areas. This work investigates an opportunity of mapping learning classifier systems (LCS) into artificial neural networks (ANN). Possible learning types for hybrid connectionist classifier system (CLCS) for multi-step problems are derived. Transformation’s opportunity of the existing LCS into CLCS is shown. Heuristic LCS and CLCS comparisons in the dynamic environment show superiority of CLCS over LCS.
Cited by 0
Olgierd Unold · Journal of Circuits Systems and Computers · 2010
This article introduces a new kind of self-adaptation in discovery mechanism of learning classifier system XCS. Unlike the previous approaches, which incorporate self-adaptive parameters in the representation of an individual, proposed model evolves competitive population of the reduced XCSs, which are able to adapt both classifiers and genetic parameters. The experimental comparisons of self-adaptive mutation rate XCS and standard XCS interacting with 11-bit, 20-bit, and 37-bit multiplexer environment were provided. It has been shown that adapting the mutation rate can give an equivalent or better performance to known good fixed parameter settings, especially for computationally complex tasks. Moreover, the self-adaptive XCS is able to solve the problem of inappropriate for a standard XCS parameters.
Cited by 10
David Pätzel, Jörg Hähner · Proceedings of the Genetic and Evolutionary Computation Conference · 2022
Learning Classifier Systems (LCSs) are a family of versatile rule-based machine learning algorithms. Despite their long research history, the foundations of most LCSs are still informal due to them having been developed in an ad-hoc manner. An exception to this is the fully Bayesian LCS described by Drugowitsch in his 2008 book. In the present paper we shortly reiterate the central points of that system and then showcase our Python implementation of it, reporting on our attempt at replicating Drugowitsch's empirical results as well as on the results of a preliminary comparison study with the well-known LCS XCSF on the same learning tasks. We are able to replicate parts of Drugowitsch's results and explain the remaining differences. The comparison results make us conclude that the system may be competitive and exhibits unique features. Finally, we identify its current greatest shortcomings and based on those the next steps towards making it more universally usable.
Cited by 2
Jason R. Wilcox, Jason R. Wilcox, Jason R. Wilcox · 1995
This thesis recasts the debate between Michigan-style and Pitt-style classifier systems to a debate on appropriately sizing organizations within a learning classifier system. Motivated by the economic study of transaction costs, an organizational classifier system (OCS) combining explicit use of multiple reputation values and organization sizing operators better distinguishes parasitic (less than optimal) classifiers than a simple classifier system (SCS). The results show that by building a system that autonomously adjusts the degree of individual to collective behavior, it is possible for it to be both efficient and resilient to problem difficulty. iii Acknowledgments I thank my parents, although no amount of thanks could sufficiently repay their unfailing love and support. I thank Dr. David Goldberg, my adviser, for his input, support, and creative ideas. I thank Kevin Carmody and Jeffrey Horn for their friendship and support. I thank Rebecca Robinson for taking care of things whil...
Cited by 0
L. Booker, W. Stolzmann, S. Wilson et al. · Virtual Community of Pathological Anatomy (University of Castilla La Mancha) · 2000
Cited by 0
Martin V. Butz, David E. Goldberg · 2002
Cited by 0
Controllers On Real, Jacob Hurst, Larry Bull et al. · 2007
To date there has only been one implementation of Holland's Learning Classifier System (LCS) on real robots. In this paper the use of Wilson's ZCS system is described for an obstacle avoidance task. Although the task is simple it does present some advances and change of emphasis over the previous LCS robotic implementation. The controller model is "event" based. Instead of the robot being assigned fixed length actions, continuous actions are taken. These actions are taken until an "event" occurs. An event can be thought of as a change of state. This division of the world into states is usually part of the problem description, and to do this automatically is currently one of the challenges facing machine learning. The paper then introduces TCS, a form of ZCS that attempts to address this issue. LCS have the ability to generalise over the state action-space. In TCS this generalisation ability can also be used to determine the extent of this space. TCS also implements components from SMDP reinforcement learning theory to weight the influence of time on the reward functions of the LCS. A simple light-seeking task on the robot platform using TCS is presented which demonstrates desirable adaptive characteristics for the use of LCS on real robots.
Cited by 0
Hai H. Dam, Hussien A. Abbass, Chris Lokan et al. · 2008
Genetics-based machine learning systems have been applied in many areas. Among them, XCS is one of the most advanced systems for reinforcement learning. UCS is a specialized case of XCS for data mining and supervised learning. In this paper we show that both XCS and UCS are quite sensitive to noise, and we propose an alternative system with Bayesian update called BCS. An experiment shows that BCS outperforms both XCS and UCS in a multiplexer problem and is quite robust to noise. I.
Cited by 0
Zoheir Ezziane · Expert Systems · 1998
In this paper, the use of genetic algorithms and classifiers systems in the graph coloring application was considered. A crossover derivative was developed to accommodate the graph coloring specifications. Also, a method for verifying and validating the new offspring coloring rules and regulations was considered. Through a series of experiments conducted on randomly generated graphs of a relatively small size, the importance of various parameters and their impact to the overall system performance is demonstrated and discussed. In conclusion, suggestions for future study regarding this topic are proposed and briefly discussed.
Cited by 0Open access
Takao Terano, Yasushi Ishikawa · 2002
This paper proposes an agent-based system with Organizational-Learning Oriented Classifier System (OCS), which is an extension of Learning Classi-fier System ~CS) into multiple agent environments. In OCS, each agent is equipped with a correspond-ing Michigan type LCS and acquires problem solv-ing knowledge based on the concepts of organizational learning in management and organizational sciences. In the proposed system, we further extend OCS to employ the following characteristics: (1) each agent solves multi-objective problems, there are some trade-offs about given problems; (2) the agents compose multi-classes, and in each class, they pursue different goals, which might cause conflicts among the classes of agents, and (3) the agents learn both individually and organizationally. We have applied the system to logical Marketing " domain in order to explain competi-tive and cooperative agent behaviors in developing and purchasing both economical nd ecological products.
Cited by 0
Rahma Ferjani, Lilia Rejeb, Chedi Abdelkarim et al. · International Journal of Information Technology & Decision Making · 2022
Learning Classifier Systems (LCSs) are a kind of evolutionary machine learning algorithms that provide highly adaptive components to deal with real world problems. They have been widely used in resolving complex problems such as decision making and classification. LCSs are flexible algorithms that are able to construct, incrementally, a set of rules and evolve them through the Evolutionary Algorithm (EA). Despite their efficiency, LCSs are not capable of handling imperfect information, which may lead to reduced performance in terms of classification accuracy. We propose a new accuracy-based Michigan-style LCS that integrates the belief function theory in the supervised classifier system. The belief function or evidence theory represents an efficient framework for treating imperfect information. The new approach shows promising results in real world classification problems.
Cited by 4
Manuel Valenzuela-Rendón · Expert Systems with Applications · 1998
Cited by 31
Petr Musı́lek, Sa Li, L. Wyard-Scott · 2005
This paper describes an enhanced learning classifier system used to evolve obstacle-avoidance rules used in mobile robot navigation. The robot learns these rules via feedback from the environment, available as sonar readings. Conventional classifiers, when used in this application, show evidence of shortcomings: becoming trapped in local minima, loss of (desirable) rules, and favouring of generalized rules. Enhancements to the classification system are described and tested using a simulated robot and environment. The enhancements prove to be worthwhile in that they overcome the limitations, and can generally handle more complex situations.
Cited by 17
Walling R. Cyre · 2003
This paper describes a modified learning classifier system for evolving grammars from small, handcrafted grammars. The algorithm uses distinct discovery and evaluation steps, and a culling step to reduce ambiguity. It performs surprisingly well in developing grammars for very complex sentences in a restricted domain of English, and in some experiments the grammar's competence improved up to tenfold.
Cited by 12
Larry Bull · Lecture notes in computer science · 2004
Cited by 7
Kreangsak Tamee, Larry Bull, Ouen Pinngern et al. · 2006
This paper presents a novel approach to clustering using a simple accuracy-based learning classifier system. Our approach achieves this by exploiting the evolutionary computing and reinforcement learning techniques inherent to such systems. The purpose of the work is to develop an approach to learning rules which accurately describe clusters without prior assumptions as to their number within a given dataset. Favourable comparisons to the commonly used k-means algorithm are demonstrated on a number of datasets
Cited by 0
Mitsuru Tanaka · ScholarWorks - UNO (University of New Orleans) · 2008
This thesis presents an implementation of a learning classifier system which learns good database schema. The system is implemented in Java using the NetBeans development environment, which provides a good control for the GUI components. The system contains four components: a user interface, a rule and message system, an apportionment of credit system, and genetic algorithms. The input of the system is a set of simple database schemas and the objective for the classifier system is to keep the good database schemas which are represented by classifiers. The learning classifier system is given some basic knowledge about database concepts or rules. The result showed that the system could decrease the bad schemas and keep the good ones.
Cited by 0Open access
K. Takadama, T. Terano, K. Shimohara et al. · Tokyo Tech Research Repository (Tokyo Institute of Technology) · 2006
Cited by 2
高玉圭樹, Keiki Takadama, 寺野隆雄 et al. · Tokyo Tech Research Repository (Tokyo Institute of Technology) · 2014
Cited by 0