Kostadin Koroutchev

Bump formation in a binary attractor neural network

The conditions for the formation of local bumps in the activity of binary attractor neural networks with spatially dependent connectivity are investigated. We show that these formations are observed when asymmetry between the activity during the retrieval and learning is imposed. An analytical approximation for the order parameters is derived. The corresponding phase diagram shows a relatively large and stable region where this effect is observed, although critical storage and information capacities drastically decrease inside that region. We demonstrate that the stability of the network, when starting from the bump formation, is larger than the stability when starting even from the whole pattern. Finally, we show a very good agreement between the analytical results and the simulations performed for different topologies of the network.

Discovering Data Set Nature through Algorithmic Clustering Based on String Compression

Text data sets can be represented using models that do not preserve text structure, or using models that preserve text structure. Our hypothesis is that depending on the data set nature, there can be advantages using a model that preserves text structure over one that does not, and vice versa. The key is to determine the best way of representing a particular data set, based on the data set itself. In this work, we proposde B“orjae to investigate this problem by combining text distortion and algorithmic clustering based on string compression. Specifically, a distortion technique previously developed by the authors is applied to destroy text structure progressively. Following this, a clustering algorithm based on string compression is used to analyze the effects of the distortion on the information contained in the texts. Several experiments are carried out on text data sets and artificially-generated data sets. The results show that in strongly structural data sets the clustering results worsen as text structure is progressively destroyed. Besides, they show that using a compressor which enables the choice of the size of the left-context symbols helps to determine the nature of the data sets. Finally, the results are contrasted with a method based on multidimensional projections and analogous conclusions are obtained.

Detecting translations of the same text and data with common source

Compression based similarity distances have the main drawback of needing the same coding scheme for the objects to be compared. In some situations, there exists significant similarity with no literal shared information: text translations, different coding schemes, etc. To overcome this problem, we present a similarity measure that compares the redundancy structure of the data extracted by means of a Lempel–Ziv compression scheme. Each text is represented as a graph in which vertices are text positions and edges represent shared information; with our measure, two texts are similar if they have the same referential topology when compressed. In this paper we give empirical evidence and a phenomenological explanation that this new measure is a robust indicator, detecting similarity between data coded in different languages. We also regard a textual data without any structure, but with a common source, and find that we can detect such data and distinguish this situation from the previous one.

Hash-Like Fractal Image Compression with Linear Execution Time

The main computational cost in Fractal Image Analysis (FIC) comes from the required range-domain full block comparisons. In this work we propose a new algorithm for this comparison, in which actual full block comparison is preceded by a very fast hash–like search of those domains close to a given range block, resulting in a performance linear with respect to the number of pixels. Once the algorithm is detailed, its results will be compared against other state–of–the–art methods in FIC.

The Social Environment as a Determinant for the Impact of the Big Five Personality Factors and the Group’s Performance

The composition of a team usually is done by having in mind the personality factors, supposing that these factors are important for the working climate and performance of the group. Starting from this hypothesis, the authors investigate the influence of the personality factors on the team achievements in a practicum of computer science students. The correlations by itself were very weak. But by partitioning the data, it results that the social environment in which the tasks are performed is a decisive factor for the importance and the influence of the team personality factors on the group’s performance. If the social environment exposes the team to situations in which the corresponding personal factors are important, then a significant correlation between these factors and the achievement is observed.

Detecting the most unusual part of a digital image

The purpose of this paper is to introduce an algorithm that can detect the most unusual part of a digital image. The most unusual part of a given shape is defined as a part of the image that has the maximal distance to all non intersecting shapes with the same form. The method can be used to scan image databases with no clear model of the interesting part or large image databases, as for example medical databases.

Functional phylogeny: The use of the sensitivity of ribosomes to protein synthesis inhibitors as a tool to study the evolution of organisms

In order to study the functional phylogeny of organisms, forty different protein synthesis inhibitors with diverse domain and functional specificities have been used to analyze forty archaeal, bacterial and eukaryotic translational systems. The inhibition curves generated with the different ribosome-antibiotic pairs have shown very interesting similarities among organisms belonging to the same phylogenetic group, confirming the feasibility of using such information in the development of evolutionary studies. A new method to extract most of the information contained in the inhibition curves is presented. Using a statistical treatment based on the principal components analysis of the data, we have defined coordinates for the organisms which have allowed us to perform a functional clustering of them. The phenograms obtained are very similar to those generated by 16/18S rRNA sequence comparison. These results prove the phylogenetic value of our functional analysis and suggest an interesting intersection between genotypic and phenotypic (functional) information.

Factorization and structure of natural 4×4 patch densities

The objective of this work is to give an approximate statistical model of the distribution of 4×4 pixel natural image patches B. Given the huge size of the sample space, we collect the required statistics not directly over B, but, instead, over a fractal compression inspired representation of B, namely by a triplet (DB,µB,B), with B being the patchs contrast, µB its brightness, and DB a codebook representation of the mean-variance normalization of B:(B?µB)/B. While not coinciding exactly with the true natural patch density p(B), the density (B)=p(DB,µB,B) should give an adequate approximation of p(B), because BBDB+µB. Our first main result is a factorization of the probability density p(D,µ,) as p(D,µ,)p(D)p(µ)p()(||B||), with being a high-contrast correction. Here, the brightness term p(µ) is largely irrelevant, and our second main result deals with the structure of the other two factors, showing that p() follows an exponential distribution, and that p(D) is uniformly distributed with respect to volume in image space. These results are largely independent of the codebook used.

Detecting the most unusual part of two- and three-dimensional digital images

The purpose of this paper is to introduce an algorithm that can detect the most unusual part of a digital image in probabilistic setting. The most unusual part of a given shape is defined as a part of the image that has the maximal distance to all non intersecting shapes with the same form. The method is tested on two and three-dimensional images and has shown very good results without any predefined model. A version of the method independent of the contrast of the image is considered and is found to be useful for finding the most unusual part (and the most similar part) of the image conditioned on given image. The results can be used to scan large image databases, as for example medical databases.

Conditions for the emergence of spatially asymmetric retrieval states in an attractor neural network

In this paper we show that during the retrieval process in a binary symmetric Hebb neural network, spatially localized states can be observed when the connectivity of the network is distance-dependent and a constraint on the activity of the network is imposed, which forces different levels of activity in the retrieval and learning states. This asymmetry in the activity during retrieval and learning is found to be a sufficient condition to observe spatially localized retrieval states. The result is confirmed analytically and by simulation.