In this paper, we examine the clonal selection algorithm CLONALG and the suggestion that it is suitable for pattern recognition. CLONALG is tested over a series of binary character recognition tasks and its performance compared to a set of basic binary matching algorithms. A number of enhancements are made to the algorithm to improve its performance and the classification tests are repeated. Re...

| We propose a new approach for leave-one-out cross-validation of neural network classiiers called \cross-validation with active pattern selection" (CV/APS). In CV/APS, the contribution of the training patterns to network learning is estimated and this information is used for active selection of CV patterns. On the tested examples, the computational cost of CV can be drastically reduced with on...

In this paper, we explain some well known experimental observations in fluid solid interaction from a thermodynamic perspective. In particular we use the extremum of the rate of entropy production to establish the stability of specific patterns observed in single and multiparticle sedimentation in an infinite fluid and the sedimentation of spheres in the presence of walls. While these phenomena...

We present an approach for encoding and specifying coherent selections in pattern hierarchies. Hierarchies of patterns of features or other modeling entities are used in architectural and mechanical CAD to eliminate laborious repetitions from the design process. Yet, often a subset or selection of the repeated occurrences in the model must be edited. Specifying a desired selection of occurrence...

SVMs tend to take a very long time to train with a large data set. If “redundant” patterns are identified and deleted in pre-processing, the training time could be reduced significantly. We propose a k-nearest neighbors(k-NN) based pattern selection method. The method tries to select the patterns that are near the decision boundary and that are correctly labeled. The simulations over synthetic ...

 Context-Aware Privacy Design Pattern Selection Siani Pearson, Yun Shen HP Laboratories HPL-2010-74 Privacy, Design Patterns, Context Awareness User-related contextual factors affect the degree of privacy protection that is necessary for a given context. Such factors include: sensitivity of data, location of data, sector, contractual restrictions, cultural expectations, user trust (in organisa...

Abatraet-We study the selection of the shape and growth velocity of three dimensionaf dendritic crystals in cubically anisotropic materials. We demonstrate that aside from minor additional complexities due to the lack of axisymmetry, the recently discovered mechanism of “microscopic solvability” can be extended to these systems and used to find a unique needle crystal solution of the equations ...

Individual classification models are recently challenged by the combined pattern recognition systems, which often show better performance. In such systems the optimal set of classifiers is first selected and then combined by a specific combination method. Large and rough search space formed from performances of various combinations of classifiers makes the selection process very difficult and o...

Fracture in quasistatically driven systems is studied by means of a discrete spring-block model. Developed from close comparison with desiccation experiments, it describes crack formation induced by friction on a substrate. The model produces cellular, hierarchical patterns of cracks, characterized by a mean fragment size linear in the layer thickness, in agreement with experiments. The selecti...

Many theoretical analyses of feedback control of pattern-forming systems assume that feedback is applied at every spatial location, something that is often difficult to accomplish in experiments. This paper considers an experimentally more feasible scenario where feedback is applied at a sparse array of discrete spatial locations. We show how such feedback can be computed analytically for a cla...