data envelopment analysis (dea) models with interval inputs and outputs have been rarely discussed in dea literature. this paper, using the enhanced russell measurement proposes an extended model which permits the presence of interval scale variables which can take both negative and positive values. the model is compared with most well-known dea models of which include the ccr model, the bcc mo...

This report deals with translation invariance of convolutional neural networks (CNNs) for automatic target recognition (ATR) from synthetic aperture radar (SAR) imagery. In particular, the translation invariance of CNNs for SAR ATR represents the robustness against misalignment of target chips extracted from SAR images. To understand the translation invariance of the CNNs, we trained CNNs which...

Super-efficiency model in the presence of negative data is a relatively neglected issue in the DEA field. The existing super-efficiency models have some shortcomings in practice. In this paper, a novel VRS radial super-efficiency DEA model based on Directional Distance Function (DDF) is proposed to provide a complete ranking order of units (including efficient and inefficient ones). The propose...

A fundamental problem in object recognition is the development of image representations that are invariant to common transformations such as translation, rotation, and small deformations. There are multiple hypotheses regarding the source of translation invariance in CNNs. One idea is that translation invariance is due to the increasing receptive field size of neurons in successive convolution ...

The general framework for quantitative technical analysis of market prices is revisited and extended. concept a global time-translation invariance its spontaneous violation restoration introduced discussed. We find that different temporal patterns leading to some famous crashes (e.g., bubbles, hockey sticks, etc.) exhibit analogous probabilistic distributions found only in the time series stock...

A CMOS implementation of a one-dimensional Rtransform with 32 inputs is presented. The implementation generates a translation invariant feature vector from a set of spatially distributed analog inputs. Repeated application to the rows and columns of an image achieves two-dimensional translation invariance. Results are presented that demonstrate the translation invariance.

Scene matching is used for image registration in many fields. There are usually translation and an arbitrary unknown rotation angle between reference and template images. The corresponding scene matching algorithm costs far more computing time than that with small rotation angle and translation. Conceptions of generalized vector image, gray-scale image rotation transformation, gray-scale image ...