Science, Art and Science Art collaborations are generally presented and understood in terms of th... more Science, Art and Science Art collaborations are generally presented and understood in terms of their products. We argue that the process of Science art can be a significant, even principal benefit of these collaborations, even though it may be largely invisible to anyone other than the collaborators. Hosting the Centenary of Canberra Science Art Commission at the Commonwealth Scientific and Industrial Research Organisation (CSIRO) has shown us that while Science and Art pursue orthogonal dimensions of creativity and innovation, collaborators can combine these directions to access new areas of imagination and ideas.
Fig. 1. MAGICal B by Eleanor Gates-Stuart from the MAGICal series (Multiparent Advanced Generatio... more Fig. 1. MAGICal B by Eleanor Gates-Stuart from the MAGICal series (Multiparent Advanced Generation Inter-Cross [17]).
Microsatellite markers have demonstrated their value for performing paternity exclusion and hence... more Microsatellite markers have demonstrated their value for performing paternity exclusion and hence exploring mating patterns in plants and animals. Methodology is well established for diploid species and several software packages exist for elucidating paternity in diploids, however these issues are not so readily addressed in polyploids due to the increased complexity of the exclusion problem and a lack of available software. We introduce PolyPatEx, an R package for paternity exclusion analysis using microsatellite data in autopolyploid, monoecious or dioecious/bisexual species with a ploidy of 4n, 6n or 8n. Given marker data for a set of offspring, their mothers, and a set of candidate fathers, PolyPatEx uses allele matching to exclude candidates whose marker alleles are incompatible with the alleles in each offspring-mother pair. PolyPatEx can analyse marker data sets in which allele copy numbers are known (genotype data) or unknown (allelic phenotype data) - for data sets in which...
In this note, a di culty with the application of Hildebrandt's closed-form training algorithm for... more In this note, a di culty with the application of Hildebrandt's closed-form training algorithm for the neocognitron is reported. In applying this algorithm we have observed that S-cells frequently fail to respond to features that they have been trained to extract. We present results which indicate that this training vector rejection in an important factor in the overall classi cation performance of the neocognitron trained using Hildebrandt's procedure.
Many kinds of arti cial neural networks have been applied to the recognition of handwritten chara... more Many kinds of arti cial neural networks have been applied to the recognition of handwritten characters. Fukushima's neocognitron is one of few networks that demonstrates invariance to input translation and tolerance of a signi cant degree of input distortion and deformation.
2011 International Conference on Digital Image Computing: Techniques and Applications, 2011
ABSTRACT Circular shortest paths represent a powerful methodology for image segmentation. The cir... more ABSTRACT Circular shortest paths represent a powerful methodology for image segmentation. The circularity condition ensures that the contour found by the algorithm is closed, a natural requirement for regular objects. Several implementations have been proposed in the past that either promise closure with high probability or ensure closure strictly, but with a mild computational efficiency handicap. Circularity can be viewed as a priori information that helps recover the correct object contour. Our "observation" is that circularity is only one among many possible constraints that can be imposed on shortest paths to guide them to a desirable solution. In this contribution, we illustrate this opportunity under a volume constraint but the concept is generally applicable. We also describe several adornments to the circular shortest path algorithm that proved useful in applications.
This paper is motivated by the need to apply compositional data analysis to molecular bioscience.... more This paper is motivated by the need to apply compositional data analysis to molecular bioscience. Awareness is gradually growing that most bioscience measurement processes yield only relative information, and that these data require appropriate analysis and interpretation. We are keen to ensure that practical compositional methods exist to provide well-principled alternatives to the potentially misleading correlation-based methods that are generally used in bioscience today.
ABSTRACT Compositional data analysis usually deals with relative information between parts where ... more ABSTRACT Compositional data analysis usually deals with relative information between parts where the total (abundances, mass, amount, etc.) is unknown or uninformative. This article addresses the question of what to do when the total is known and is of interest. Tools used in this case are reviewed and analysed, in particular the relationship between the positive orthant of D-dimensional real space, the product space of the real line times the D-part simplex, and their Euclidean space structures. The first alternative corresponds to data analysis taking logarithms on each component, and the second one to treat a log-transformed total jointly with a composition describing the distribution of component amounts. Real data about total abundances of phytoplankton in an Australian river motivated the present study and are used for illustration.
ABSTRACT Compositional data analysis methods aim to help us understand data that carry only relat... more ABSTRACT Compositional data analysis methods aim to help us understand data that carry only relative information. They can be used when information about totals (e.g., size, mass, amount, or absolute abundance) is unknown, or considered uninformative. This paper addresses a situation where both composition and size are known, and both are potentially informative. Within this framework we analyse jointly composition and size on the total abundances of phytoplankton in a river.
A difficulty with the application of the closed-form training algorithm for the neocognitron prop... more A difficulty with the application of the closed-form training algorithm for the neocognitron proposed by T.H. Hildebrandt (ibid., vol.2, p.557-88, Nov. 1991) is reported. In applying this algorithm the commenters have observed that S-cells frequently fail to respond to features that they have been trained to extract. Results which indicate that this training vector rejection in an important factor in the overall classification performance of the neocognitron trained using Hildebrandt's procedure are presented. In reply, Hildebrandt explains that the negative results obtained by the commenter are not specific to the proposed algorithm and are easily explained in terms of set theory.
We describe a sequence of experiments investigating the strengths and limitations of Fukushima's ... more We describe a sequence of experiments investigating the strengths and limitations of Fukushima's neocognitron as a handwritten digit classifier. Using the results of these experiments as a foundation, we propose and evaluate improvements to Fukushima's original network in an effort to obtain higher recognition performance. The neocognitron's performance is shown to be strongly dependent on the choice of selectivity parameters and we present two methods to adjust these variables. Performance of the network under the more effective of the two new selectivity adjustment techniques suggests that the network fails to exploit the features that distinguish different classes of input data. To avoid this shortcoming, the network's final layer cells were replaced by a nonlinear classifier (a multilayer perceptron) to create a hybrid architecture. Tests of Fukushima's original system and the novel systems proposed in this paper suggest that it may be difficult for the neocognitron to achieve the performance of existing digit classifiers due to its reliance upon the supervisor's choice of selectivity parameters and training data. These findings pertain to Fukushima's implementation of the system and should not be seen as diminishing the practical significance of the concept of hierarchical feature extraction embodied in the neocognitron.
... The likelihood expression for an observation given the current state was modi ed by the addit... more ... The likelihood expression for an observation given the current state was modi ed by the addition of a term describing the probability of the state having a given energy. ... 2 4 6 8 Figure 3: Clean Speech (\Add Yankee") sound A0775S01.WAV] Time (ms) Frequency (kHz) ...
Science, Art and Science Art collaborations are generally presented and understood in terms of th... more Science, Art and Science Art collaborations are generally presented and understood in terms of their products. We argue that the process of Science art can be a significant, even principal benefit of these collaborations, even though it may be largely invisible to anyone other than the collaborators. Hosting the Centenary of Canberra Science Art Commission at the Commonwealth Scientific and Industrial Research Organisation (CSIRO) has shown us that while Science and Art pursue orthogonal dimensions of creativity and innovation, collaborators can combine these directions to access new areas of imagination and ideas.
Fig. 1. MAGICal B by Eleanor Gates-Stuart from the MAGICal series (Multiparent Advanced Generatio... more Fig. 1. MAGICal B by Eleanor Gates-Stuart from the MAGICal series (Multiparent Advanced Generation Inter-Cross [17]).
Microsatellite markers have demonstrated their value for performing paternity exclusion and hence... more Microsatellite markers have demonstrated their value for performing paternity exclusion and hence exploring mating patterns in plants and animals. Methodology is well established for diploid species and several software packages exist for elucidating paternity in diploids, however these issues are not so readily addressed in polyploids due to the increased complexity of the exclusion problem and a lack of available software. We introduce PolyPatEx, an R package for paternity exclusion analysis using microsatellite data in autopolyploid, monoecious or dioecious/bisexual species with a ploidy of 4n, 6n or 8n. Given marker data for a set of offspring, their mothers, and a set of candidate fathers, PolyPatEx uses allele matching to exclude candidates whose marker alleles are incompatible with the alleles in each offspring-mother pair. PolyPatEx can analyse marker data sets in which allele copy numbers are known (genotype data) or unknown (allelic phenotype data) - for data sets in which...
In this note, a di culty with the application of Hildebrandt's closed-form training algorithm for... more In this note, a di culty with the application of Hildebrandt's closed-form training algorithm for the neocognitron is reported. In applying this algorithm we have observed that S-cells frequently fail to respond to features that they have been trained to extract. We present results which indicate that this training vector rejection in an important factor in the overall classi cation performance of the neocognitron trained using Hildebrandt's procedure.
Many kinds of arti cial neural networks have been applied to the recognition of handwritten chara... more Many kinds of arti cial neural networks have been applied to the recognition of handwritten characters. Fukushima's neocognitron is one of few networks that demonstrates invariance to input translation and tolerance of a signi cant degree of input distortion and deformation.
2011 International Conference on Digital Image Computing: Techniques and Applications, 2011
ABSTRACT Circular shortest paths represent a powerful methodology for image segmentation. The cir... more ABSTRACT Circular shortest paths represent a powerful methodology for image segmentation. The circularity condition ensures that the contour found by the algorithm is closed, a natural requirement for regular objects. Several implementations have been proposed in the past that either promise closure with high probability or ensure closure strictly, but with a mild computational efficiency handicap. Circularity can be viewed as a priori information that helps recover the correct object contour. Our "observation" is that circularity is only one among many possible constraints that can be imposed on shortest paths to guide them to a desirable solution. In this contribution, we illustrate this opportunity under a volume constraint but the concept is generally applicable. We also describe several adornments to the circular shortest path algorithm that proved useful in applications.
This paper is motivated by the need to apply compositional data analysis to molecular bioscience.... more This paper is motivated by the need to apply compositional data analysis to molecular bioscience. Awareness is gradually growing that most bioscience measurement processes yield only relative information, and that these data require appropriate analysis and interpretation. We are keen to ensure that practical compositional methods exist to provide well-principled alternatives to the potentially misleading correlation-based methods that are generally used in bioscience today.
ABSTRACT Compositional data analysis usually deals with relative information between parts where ... more ABSTRACT Compositional data analysis usually deals with relative information between parts where the total (abundances, mass, amount, etc.) is unknown or uninformative. This article addresses the question of what to do when the total is known and is of interest. Tools used in this case are reviewed and analysed, in particular the relationship between the positive orthant of D-dimensional real space, the product space of the real line times the D-part simplex, and their Euclidean space structures. The first alternative corresponds to data analysis taking logarithms on each component, and the second one to treat a log-transformed total jointly with a composition describing the distribution of component amounts. Real data about total abundances of phytoplankton in an Australian river motivated the present study and are used for illustration.
ABSTRACT Compositional data analysis methods aim to help us understand data that carry only relat... more ABSTRACT Compositional data analysis methods aim to help us understand data that carry only relative information. They can be used when information about totals (e.g., size, mass, amount, or absolute abundance) is unknown, or considered uninformative. This paper addresses a situation where both composition and size are known, and both are potentially informative. Within this framework we analyse jointly composition and size on the total abundances of phytoplankton in a river.
A difficulty with the application of the closed-form training algorithm for the neocognitron prop... more A difficulty with the application of the closed-form training algorithm for the neocognitron proposed by T.H. Hildebrandt (ibid., vol.2, p.557-88, Nov. 1991) is reported. In applying this algorithm the commenters have observed that S-cells frequently fail to respond to features that they have been trained to extract. Results which indicate that this training vector rejection in an important factor in the overall classification performance of the neocognitron trained using Hildebrandt's procedure are presented. In reply, Hildebrandt explains that the negative results obtained by the commenter are not specific to the proposed algorithm and are easily explained in terms of set theory.
We describe a sequence of experiments investigating the strengths and limitations of Fukushima's ... more We describe a sequence of experiments investigating the strengths and limitations of Fukushima's neocognitron as a handwritten digit classifier. Using the results of these experiments as a foundation, we propose and evaluate improvements to Fukushima's original network in an effort to obtain higher recognition performance. The neocognitron's performance is shown to be strongly dependent on the choice of selectivity parameters and we present two methods to adjust these variables. Performance of the network under the more effective of the two new selectivity adjustment techniques suggests that the network fails to exploit the features that distinguish different classes of input data. To avoid this shortcoming, the network's final layer cells were replaced by a nonlinear classifier (a multilayer perceptron) to create a hybrid architecture. Tests of Fukushima's original system and the novel systems proposed in this paper suggest that it may be difficult for the neocognitron to achieve the performance of existing digit classifiers due to its reliance upon the supervisor's choice of selectivity parameters and training data. These findings pertain to Fukushima's implementation of the system and should not be seen as diminishing the practical significance of the concept of hierarchical feature extraction embodied in the neocognitron.
... The likelihood expression for an observation given the current state was modi ed by the addit... more ... The likelihood expression for an observation given the current state was modi ed by the addition of a term describing the probability of the state having a given energy. ... 2 4 6 8 Figure 3: Clean Speech (\Add Yankee") sound A0775S01.WAV] Time (ms) Frequency (kHz) ...
Uploads
Papers by David Lovell