Edmund Ronald - Academia.edu (original) (raw)
Papers by Edmund Ronald
The control problem of soft-landing a toy lunar module simulation is investigated in the context ... more The control problem of soft-landing a toy lunar module simulation is investigated in the context of neural nets. While traditional supervised back-propagation training is inappropriate for lack of training exemplars, genetic algorithms allow a controller to be evolved without di culty: Evolution is a form of unsupervised learning. A novelty introduced in this paper is the presentation of additional renormalized inputs to the net; experiments indicate that the presence of such inputs allows precision of control to be attained faster, when learning time is measured by the number of generations for which the GA must run to attain a certain mean performance.
MILLION MODULE NEURAL SYSTEMS EVOLUTION - The Next Step in ATR's Billion Neuron Artificial Brain ("CAM-Brain") Project
This position paper discusses the evolution of multi-module neural net systems, where the number ... more This position paper discusses the evolution of multi-module neural net systems, where the number of neural net modules is up to ten million (i.e. an artificial brain). ATR's CAM-Brain Project [de Garis 1993, 1996] has progressed to the point where it is technically possible (using a new FPGA (Field Programmable Gate Array) based evolvable hardware (EHW or E-Hard) system to
Artificial Evolution
Evolutionary algorithms (EAs) have been increasingly, and successfully, applied to combinatorial ... more Evolutionary algorithms (EAs) have been increasingly, and successfully, applied to combinatorial optimization problems. However, EAs are relatively complicated algorithms (compared to local search, for example) and it is not always clear to what extent their behaviour can be explained by the particular set of strategies and parameters used. One of the most commonly-used metaphors to describe the process of simple methods such as local search is that of a ‘fitness landscape’, but even in this case, describing what we mean by such a term is not as easy as might be assumed. In this paper, we first present some intuitive ideas and mathematical definitions of what is meant by a landscape and its properties, and review some of the theoretical and experimental work that has been carried out over the past 6 years. We then consider how the concepts associated with a landscape can be extended to search by means of evolutionary algorithms, and connect this with previous work on epistasis varia...
International Journal of Modern Physics C, 1998
We show that there exists a simple solution to the density problem in cellular automata, under fi... more We show that there exists a simple solution to the density problem in cellular automata, under fixed boundary conditions, in contrast to previously used periodic ones.
Lecture Notes in Computer Science, 1994
The control problem of soft-landing a toy lunar module simulation is investigated in the context ... more The control problem of soft-landing a toy lunar module simulation is investigated in the context of neural nets. While traditional supervised back-propagation training is inappropriate for lack of training exemplars, genetic algorithms allow a controller to be evolved without di culty: Evolution is a form of unsupervised learning. A novelty introduced in this paper is the presentation of additional renormalized inputs to the net; experiments indicate that the presence of such inputs allows precision of control to be attained faster, when learning time is measured by the number of generations for which the GA must run to attain a certain mean performance.
Robotics and Autonomous Systems, 2001
We examine the eventual role of surprise in two domains of human endeavor: classical engineering ... more We examine the eventual role of surprise in two domains of human endeavor: classical engineering and what we call "emergent engineering", with examples relevant to the field of robotics. Placing ourselves within the formal framework of the recently proposed "emergence test", we argue that the element of surprise, central in the test, serves to illuminate a fundamental difference between industrial and autonomous robots: unsurprise is demanded of classically engineered automation, while a mild form of surprise -unsurprising surprise -must of necessity be tolerated in biologically inspired systems, including behavior-based robotics.
A new species of hardware
IEEE Spectrum, 2000
... A new species of hardware MOSHF SIPPER Swiss Federal Institute of Toch I101 og y d EDMUND M .... more ... A new species of hardware MOSHF SIPPER Swiss Federal Institute of Toch I101 og y d EDMUND M . A. RONALD E CO I c Po I y te c h ri i yu e , Ccnter for Applied Mat hema tics Page 2. Evolved Circuit RFB le6 i 1 VSRC % 'i II : RLDAD 10 I I I 1 i - .. . ._ ...
Design, Observation, Surprise! A Test of Emergence
Artificial Life, 1999
The field of artificial life (Alife) is replete with documented instances of emergence, though de... more The field of artificial life (Alife) is replete with documented instances of emergence, though debate still persists as to the meaning of this term. We contend that, in the absence of an acceptable definition, researchers in the field would be well served by adopting an emergence certification mark that would garner approval from the Alife community. Toward this end, we propose an emergence test, namely, criteria by which one can justify conferring the emergence label.
Genetic Extensions of Neural Net Learning: Transfer Functions and Renormalisation Coefficients
This paper deals with technical issues relevant to artificial neural net (ANN) training by geneti... more This paper deals with technical issues relevant to artificial neural net (ANN) training by genetic algorithms. Neural nets have applications ranging from perception to control; in the context of control, achieving great precision is more critical than in pattern recognition or classification tasks. In previous work, the authors have found that when employing genetic search to train a net, both precision and training speed can be greatly enhanced by an input renormalisation technique. In this paper we investigate the automatic tuning of such renormalisation coefficients, as well as the tuning of the slopes of the transfer functions of the individual neurons in the net. Waiting time analysis is presented as an alternative to the classical "mean perfomance" interpretation of GA experiments. It is felt that it provides a more realistic evaluation of the real-world usefulness of a GA.
Lecture Notes in Computer Science, 1999
The field of artificial life (Alife) is replete with documented instances of emergence, though de... more The field of artificial life (Alife) is replete with documented instances of emergence, though debate still persists as to the meaning of this term. In the absence of a formal definition, researchers in the field would be well served by adopting an emergence certification maxk which would garner approval from the Alife community. We propose an emergence test, consisting of three criteria-design, observation, and surprise-for conferring the emergence label.
Neuro-genetic Truck Backer-upper Controller I. Plant Description
Marc Schoenauer and Edmund Ronald Abstract|The precise docking of a truck at a loading dock has b... more Marc Schoenauer and Edmund Ronald Abstract|The precise docking of a truck at a loading dock has been proposed in [Nguyen & Widrow 90] as a benchmark problem for non-linear control by neural-nets. The main di culty is that back-propagation is not a priori suitable as a learning paradigm, because no set of training vectors is available: It is non-trivial to nd solution trajectories that dock the truck from anywhere in the loading yard. In this paper we show how a genetic algorithm can evolve the weights of a feedforward 3-layer neural net that solves the control problem for a given starting state, achieving a short trajectory from starting point to goal. The tness of a net in the population is a function of both the nearest position from the goal and the distance travelled. The in uence of input data renormalisation on trajectory precision is also discussed. I. PLANT DESCRIPTION The plant to be controlled here is a truck with trailer, which should be backed up to a loading dock; the t...
When Selection Meets Seduction
Proceedings of the 6th International Conference on Genetic Algorithms, 1995
Apprentissage evolutionniste des reseaux neuromimetiques
Http Www Theses Fr, 1997
Neuro-Genetic Truck Backer-Upper Controller
The precise docking of a truck at a loading dockhas been proposed in [Nguyen & Wi... more The precise docking of a truck at a loading dockhas been proposed in [Nguyen & Widrow 90] as a benchmarkproblem for non-linear control by neural-nets. Themain difficulty is that back-propagation is not a priori suitableas a learning paradigm, because no set of training vectorsis available: It is non-trivial to find solution trajectoriesthat dock the truck from anywhere in the loading yard.In this paper we show how a genetic algorithm can evolvethe weights of a feedforward 3-layer neural ...
Feature induction by backpropagation
Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94), 1994
A method for investigating the internal knowledge representation constructed by neural net learni... more A method for investigating the internal knowledge representation constructed by neural net learning is described: it is shown how from a given weight matrix defining a feedforward artificial neural net, we can induce characteristic patterns of each of the classes of inputs classified by that net. These characteristic patterns, called prototypes, are found by a gradient descent search of the space of inputs. After an exposition of the theory, results are given for the well known LED recognition problem where a network simulates recognition of decimal digits displayed on a seven-segment LED display. Contrary to theoretical intuition, the experimental results indicate that the computed prototypes retain only some of the features of the original input patterns. Thus it appears that the indicated method extracts those features deemed significant by the net
Neuro-genetic truck backer-upper controller
Proceedings of the First IEEE Conference on Evolutionary Computation. IEEE World Congress on Computational Intelligence, 1994
The precise docking of a truck at a loading dockhas been proposed in [Nguyen & Wi... more The precise docking of a truck at a loading dockhas been proposed in [Nguyen & Widrow 90] as a benchmarkproblem for non-linear control by neural-nets. Themain difficulty is that back-propagation is not a priori suitableas a learning paradigm, because no set of training vectorsis available: It is non-trivial to find solution trajectoriesthat dock the truck from anywhere in the loading yard.In this paper we show how a genetic algorithm can evolvethe weights of a feedforward 3-layer neural ...
Artificial Evolution, 1997
This position paper discusses the evolution of multi-module neural net systems, where the number ... more This position paper discusses the evolution of multi-module neural net systems, where the number of neural net modules is up to ten million (i.e. an "artificial brain"). ATR's "CAM-Brain" Project [de Garis 1993, 1996] has progressed to the point where it is technically possible (using a new FPGA (Field Programmable Gate Array) based evolvable hardware (EHW or E-Hard) system to be completed by the spring of 1998 [Korkin & de Garis 1997]) to begin to evolve and build an artificial brain containing 10,000 neural net modules. This development raises the prospect that within a few years these numbers will rapidly increase. This paper introduces some issues that such massive systembuilding will generate. The immediate question is "What should we evolve?" This paper presents some suggested evolvable system targets containing N neural net modules, where N =
Artificial Evolution
Lecture Notes in Computer Science, 2000
Artificial Evolution
The Journal of Artificial Societies and Social Simulation, 2002
The control problem of soft-landing a toy lunar module simulation is investigated in the context ... more The control problem of soft-landing a toy lunar module simulation is investigated in the context of neural nets. While traditional supervised back-propagation training is inappropriate for lack of training exemplars, genetic algorithms allow a controller to be evolved without di culty: Evolution is a form of unsupervised learning. A novelty introduced in this paper is the presentation of additional renormalized inputs to the net; experiments indicate that the presence of such inputs allows precision of control to be attained faster, when learning time is measured by the number of generations for which the GA must run to attain a certain mean performance.
MILLION MODULE NEURAL SYSTEMS EVOLUTION - The Next Step in ATR's Billion Neuron Artificial Brain ("CAM-Brain") Project
This position paper discusses the evolution of multi-module neural net systems, where the number ... more This position paper discusses the evolution of multi-module neural net systems, where the number of neural net modules is up to ten million (i.e. an artificial brain). ATR's CAM-Brain Project [de Garis 1993, 1996] has progressed to the point where it is technically possible (using a new FPGA (Field Programmable Gate Array) based evolvable hardware (EHW or E-Hard) system to
Artificial Evolution
Evolutionary algorithms (EAs) have been increasingly, and successfully, applied to combinatorial ... more Evolutionary algorithms (EAs) have been increasingly, and successfully, applied to combinatorial optimization problems. However, EAs are relatively complicated algorithms (compared to local search, for example) and it is not always clear to what extent their behaviour can be explained by the particular set of strategies and parameters used. One of the most commonly-used metaphors to describe the process of simple methods such as local search is that of a ‘fitness landscape’, but even in this case, describing what we mean by such a term is not as easy as might be assumed. In this paper, we first present some intuitive ideas and mathematical definitions of what is meant by a landscape and its properties, and review some of the theoretical and experimental work that has been carried out over the past 6 years. We then consider how the concepts associated with a landscape can be extended to search by means of evolutionary algorithms, and connect this with previous work on epistasis varia...
International Journal of Modern Physics C, 1998
We show that there exists a simple solution to the density problem in cellular automata, under fi... more We show that there exists a simple solution to the density problem in cellular automata, under fixed boundary conditions, in contrast to previously used periodic ones.
Lecture Notes in Computer Science, 1994
The control problem of soft-landing a toy lunar module simulation is investigated in the context ... more The control problem of soft-landing a toy lunar module simulation is investigated in the context of neural nets. While traditional supervised back-propagation training is inappropriate for lack of training exemplars, genetic algorithms allow a controller to be evolved without di culty: Evolution is a form of unsupervised learning. A novelty introduced in this paper is the presentation of additional renormalized inputs to the net; experiments indicate that the presence of such inputs allows precision of control to be attained faster, when learning time is measured by the number of generations for which the GA must run to attain a certain mean performance.
Robotics and Autonomous Systems, 2001
We examine the eventual role of surprise in two domains of human endeavor: classical engineering ... more We examine the eventual role of surprise in two domains of human endeavor: classical engineering and what we call "emergent engineering", with examples relevant to the field of robotics. Placing ourselves within the formal framework of the recently proposed "emergence test", we argue that the element of surprise, central in the test, serves to illuminate a fundamental difference between industrial and autonomous robots: unsurprise is demanded of classically engineered automation, while a mild form of surprise -unsurprising surprise -must of necessity be tolerated in biologically inspired systems, including behavior-based robotics.
A new species of hardware
IEEE Spectrum, 2000
... A new species of hardware MOSHF SIPPER Swiss Federal Institute of Toch I101 og y d EDMUND M .... more ... A new species of hardware MOSHF SIPPER Swiss Federal Institute of Toch I101 og y d EDMUND M . A. RONALD E CO I c Po I y te c h ri i yu e , Ccnter for Applied Mat hema tics Page 2. Evolved Circuit RFB le6 i 1 VSRC % 'i II : RLDAD 10 I I I 1 i - .. . ._ ...
Design, Observation, Surprise! A Test of Emergence
Artificial Life, 1999
The field of artificial life (Alife) is replete with documented instances of emergence, though de... more The field of artificial life (Alife) is replete with documented instances of emergence, though debate still persists as to the meaning of this term. We contend that, in the absence of an acceptable definition, researchers in the field would be well served by adopting an emergence certification mark that would garner approval from the Alife community. Toward this end, we propose an emergence test, namely, criteria by which one can justify conferring the emergence label.
Genetic Extensions of Neural Net Learning: Transfer Functions and Renormalisation Coefficients
This paper deals with technical issues relevant to artificial neural net (ANN) training by geneti... more This paper deals with technical issues relevant to artificial neural net (ANN) training by genetic algorithms. Neural nets have applications ranging from perception to control; in the context of control, achieving great precision is more critical than in pattern recognition or classification tasks. In previous work, the authors have found that when employing genetic search to train a net, both precision and training speed can be greatly enhanced by an input renormalisation technique. In this paper we investigate the automatic tuning of such renormalisation coefficients, as well as the tuning of the slopes of the transfer functions of the individual neurons in the net. Waiting time analysis is presented as an alternative to the classical "mean perfomance" interpretation of GA experiments. It is felt that it provides a more realistic evaluation of the real-world usefulness of a GA.
Lecture Notes in Computer Science, 1999
The field of artificial life (Alife) is replete with documented instances of emergence, though de... more The field of artificial life (Alife) is replete with documented instances of emergence, though debate still persists as to the meaning of this term. In the absence of a formal definition, researchers in the field would be well served by adopting an emergence certification maxk which would garner approval from the Alife community. We propose an emergence test, consisting of three criteria-design, observation, and surprise-for conferring the emergence label.
Neuro-genetic Truck Backer-upper Controller I. Plant Description
Marc Schoenauer and Edmund Ronald Abstract|The precise docking of a truck at a loading dock has b... more Marc Schoenauer and Edmund Ronald Abstract|The precise docking of a truck at a loading dock has been proposed in [Nguyen & Widrow 90] as a benchmark problem for non-linear control by neural-nets. The main di culty is that back-propagation is not a priori suitable as a learning paradigm, because no set of training vectors is available: It is non-trivial to nd solution trajectories that dock the truck from anywhere in the loading yard. In this paper we show how a genetic algorithm can evolve the weights of a feedforward 3-layer neural net that solves the control problem for a given starting state, achieving a short trajectory from starting point to goal. The tness of a net in the population is a function of both the nearest position from the goal and the distance travelled. The in uence of input data renormalisation on trajectory precision is also discussed. I. PLANT DESCRIPTION The plant to be controlled here is a truck with trailer, which should be backed up to a loading dock; the t...
When Selection Meets Seduction
Proceedings of the 6th International Conference on Genetic Algorithms, 1995
Apprentissage evolutionniste des reseaux neuromimetiques
Http Www Theses Fr, 1997
Neuro-Genetic Truck Backer-Upper Controller
The precise docking of a truck at a loading dockhas been proposed in [Nguyen & Wi... more The precise docking of a truck at a loading dockhas been proposed in [Nguyen & Widrow 90] as a benchmarkproblem for non-linear control by neural-nets. Themain difficulty is that back-propagation is not a priori suitableas a learning paradigm, because no set of training vectorsis available: It is non-trivial to find solution trajectoriesthat dock the truck from anywhere in the loading yard.In this paper we show how a genetic algorithm can evolvethe weights of a feedforward 3-layer neural ...
Feature induction by backpropagation
Proceedings of 1994 IEEE International Conference on Neural Networks (ICNN'94), 1994
A method for investigating the internal knowledge representation constructed by neural net learni... more A method for investigating the internal knowledge representation constructed by neural net learning is described: it is shown how from a given weight matrix defining a feedforward artificial neural net, we can induce characteristic patterns of each of the classes of inputs classified by that net. These characteristic patterns, called prototypes, are found by a gradient descent search of the space of inputs. After an exposition of the theory, results are given for the well known LED recognition problem where a network simulates recognition of decimal digits displayed on a seven-segment LED display. Contrary to theoretical intuition, the experimental results indicate that the computed prototypes retain only some of the features of the original input patterns. Thus it appears that the indicated method extracts those features deemed significant by the net
Neuro-genetic truck backer-upper controller
Proceedings of the First IEEE Conference on Evolutionary Computation. IEEE World Congress on Computational Intelligence, 1994
The precise docking of a truck at a loading dockhas been proposed in [Nguyen & Wi... more The precise docking of a truck at a loading dockhas been proposed in [Nguyen & Widrow 90] as a benchmarkproblem for non-linear control by neural-nets. Themain difficulty is that back-propagation is not a priori suitableas a learning paradigm, because no set of training vectorsis available: It is non-trivial to find solution trajectoriesthat dock the truck from anywhere in the loading yard.In this paper we show how a genetic algorithm can evolvethe weights of a feedforward 3-layer neural ...
Artificial Evolution, 1997
This position paper discusses the evolution of multi-module neural net systems, where the number ... more This position paper discusses the evolution of multi-module neural net systems, where the number of neural net modules is up to ten million (i.e. an "artificial brain"). ATR's "CAM-Brain" Project [de Garis 1993, 1996] has progressed to the point where it is technically possible (using a new FPGA (Field Programmable Gate Array) based evolvable hardware (EHW or E-Hard) system to be completed by the spring of 1998 [Korkin & de Garis 1997]) to begin to evolve and build an artificial brain containing 10,000 neural net modules. This development raises the prospect that within a few years these numbers will rapidly increase. This paper introduces some issues that such massive systembuilding will generate. The immediate question is "What should we evolve?" This paper presents some suggested evolvable system targets containing N neural net modules, where N =
Artificial Evolution
Lecture Notes in Computer Science, 2000
Artificial Evolution
The Journal of Artificial Societies and Social Simulation, 2002