Towards finding optimal mixture of subspaces for data classification

Musa, Mohamed Elhafiz Mustafa
In pattern recognition, when data has different structures in different parts of the input space, fitting one global model can be slow and inaccurate. Learning methods can quickly learn the structure of the data in local regions, consequently, offering faster and more accurate model fitting. Breaking training data set into smaller subsets may lead to curse of dimensionality problem, as a training sample subset may not be enough for estimating the required set of parameters for the submodels. Increasing the size of training data may not be at hand in many situations. Interestingly, the data in local regions becomes more correlated. Therefore, by decorrelation methods we can reduce data dimensions and hence the number of parameters. In other words, we can find uncorrelated low dimensional subspaces that capture most of the data variability. The current subspace modelling methods have proved better performance than the global modelling methods for the given type of training data structure. Nevertheless these methods still need more research work as they are suffering from two limitations 2 There is no standard method to specify the optimal number of subspaces. ² There is no standard method to specify the optimal dimensionality for each subspace. In the current models these two parameters are determined beforehand. In this dissertation we propose and test algorithms that try to find a suboptimal number of principal subspaces and a suboptimal dimensionality for each principal subspaces automatically.


Continuous dimensionality characterization of image structures
Felsberg, Michael; Kalkan, Sinan; Kruger, Norbert (Elsevier BV, 2009-05-04)
Intrinsic dimensionality is a concept introduced by statistics and later used in image processing to measure the dimensionality of a data set. In this paper, we introduce a continuous representation of the intrinsic dimension of an image patch in terms of its local spectrum or, equivalently, its gradient field. By making use of a cone structure and barycentric co-ordinates, we can associate three confidences to the three different ideal cases of intrinsic dimensions corresponding to homogeneous image patche...
A Comparative Study on Distance Metrics in Self- Supervised Unstructured Road Detection Domain
Özütemiz, Kadri Buğra; Hacınecipoğlu, Akif; Koku, Ahmet Buğra; Konukseven, Erhan İlhan (2013-09-20)
In pattern recognition/machine learning domain, selecting appropriate distance metric for the problem to find the distance between feature vectors or the distance between a feature vector and decision boundary is important in order to have satisfying results from the algorithm designed. In this study, in order to find the most appropriate distance metric to use in classification of road/non-road regions in streaming images, 6 different distance metrics are implemented and their classification performances a...
A Formal Methods Approach to Pattern Recognition and Synthesis in Reaction Diffusion Networks
Bartocci, Ezio; Aydın Göl, Ebru; Haghighi, Iman; Belta, Calin (2018-03-01)
We introduce a formal framework for specifying, detecting, and generating spatial patterns in reaction diffusion networks. Our approach is based on a novel spatial superposition logic, whose semantics is defined over the quad-tree representation of a partitioned image. We demonstrate how to use rule-based classifiers to efficiently learn spatial superposition logic formulas for several types of patterns from positive and negative examples. We implement pattern detection as a model-checking algorithm and we ...
Dimension reduction using global and local pattern information-based maximum margin criterion
Sakarya, Ufuk (2016-07-01)
Dimension reduction is an important research area in pattern recognition when dealing with high-dimensional data. In this paper, a novel supervised dimension reduction approach is introduced for classification. Advantages of using not only global pattern information but also local pattern information are examined in the maximum margin criterion framework. Experimental comparative results in object recognition, handwritten digit recognition, and hyperspectral image classification are presented. According to ...
Low-level multiscale image segmentation and a benchmark for its evaluation
Akbaş, Emre (Elsevier BV, 2020-10-01)
In this paper, we present a segmentation algorithm to detect low-level structure present in images. The algorithm is designed to partition a given image into regions, corresponding to image structures, regardless of their shapes, sizes, and levels of interior homogeneity. We model a region as a connected set of pixels that is surrounded by ramp edge discontinuities where the magnitude of these discontinuities is large compared to the variation inside the region. Each region is associated with a scale that d...
Citation Formats
M. E. M. Musa, “Towards finding optimal mixture of subspaces for data classification,” Ph.D. - Doctoral Program, Middle East Technical University, 2003.