Stefan Hadjitodorov | Bulgarian Academy of Sciences (original) (raw)

Uploads

Papers by Stefan Hadjitodorov

Inženerni nauki, Jun 3, 2024

Critical infrastructure constitutes the backbone of modern society, encompassing sectors such as ... more Critical infrastructure constitutes the backbone of modern society, encompassing sectors such as energy, transportation, and water supply. Risk assessment of critical infrastructure objects is an important task that helps to identify and analyse the potential threats and vulnerabilities of such objects. In recent years, especially after the outbreak of the war in Ukraine, the importance of space infrastructures (ground and space) has grown and they were included in the list of 11 areas of critical infrastructure in the latest EU directive. On the other hand, space technologies appear in key enablers to support resilient & sustainable Critical Infrastructure. This study examines the novel combination of multispectral satellite images and powerful machine learning algorithms, which together have exceptional prospects for enhancing the effectiveness and precision of evaluating and overseeing crucial infrastructure systems.

Electronics Letters, 1991

Cluster ensembles are deemed to be better than single clustering algorithms for discovering compl... more Cluster ensembles are deemed to be better than single clustering algorithms for discovering complex or noisy structures in data. We consider difierent heuristics to introduce diversity in cluster ensembles and study their individual and combined efiect on the ensemble accuracy. Our experiments with three artiflcial and three real data sets, and 12 ensemble types, showed that the most successful diversifying

Electronics Letters, 1993

... The data path itself occupies 650000 transistors and operates at a projected clock speed of 4... more ... The data path itself occupies 650000 transistors and operates at a projected clock speed of 40MHz (devices due back during October 1993 ... by means of the short-term autocorrelation function R,,(T). To increase the accuracy the signal is preprocessed by central clipping [I]. The ...

Comptes Rendus De L Academie Bulgare Des Sciences, 2003

ABSTRACT In this paper a generalized nets model of an access permission system based on the evalu... more ABSTRACT In this paper a generalized nets model of an access permission system based on the evalution of biometric parametes is presented. Generalized nets are defined as extensions of the ordinary Petri nets in the sense that they give more and greater modelling possibilities. This is the first example of application of generalized nets models with separate subnets describing such a system.

Springer eBooks, 2021

The research and results in the field of acoustic analysis of voices for different purposes—laryn... more The research and results in the field of acoustic analysis of voices for different purposes—laryngeal pathology detection, speaker recognition and evaluation of the emotional state at the Center on Biomedical Engineering, Bulgarian Academy of Sciences (now Institute of Biophysics and Biomedical Engineering) has been described. A lot of novelties have been proposed and introduced regarding: acoustical parameters and approaches; methods for evaluation of well-known acoustical parameters; methods for classification and recognition; computer (software) systems for acoustic analysis of voices and speaker recognition.

DOAJ (DOAJ: Directory of Open Access Journals), Oct 1, 2007

The article focuses on a type of client/server multiagent system. Such system consists of several... more The article focuses on a type of client/server multiagent system. Such system consists of several server environments, each populated by a single server agent and multiple client agents. We propose that a server agent should monitor clients activity in order to perform better. We employ the term agent tracking, introduced by Tambe and Rosenbloom. We introduce the idea of distributed

The Ceneralized Nets (GN) that are an extension of Petri Nets have been used for modeling of di®e... more The Ceneralized Nets (GN) that are an extension of Petri Nets have been used for modeling of di®erent processes, including intellectual ones. In this paper a CN model is developed for the description of the functioning of an biometrics based access-control system. The model gives the possibility for the investigation and practical realization of speci¯c systems.

European Society for Fuzzy Logic and Technology Conference, 2003

Pattern Recognition Letters, 1990

ABSTRACT The term ‘key features’ is introduced to denote features with rare occurrence and great ... more ABSTRACT The term ‘key features’ is introduced to denote features with rare occurrence and great diagnostic power, which are implemented in a two-level classification scheme. Experiments with medical data illustrate the idea, confirming the expected improvement of the classification accuracy.

Fuzzy Sets and Systems, Aug 1, 1992

ABSTRACT A version for increasing of the classification accuracy of one of the basic statistical ... more ABSTRACT A version for increasing of the classification accuracy of one of the basic statistical nonparametrical methods, the K-NN method, is proposed. Here the idea of including fuzzy information in the K-NN method is applied in a new way. The distances are modified by means of the pattern degrees of membership to the classes to which the pattern belongs. Thus for each of the labeled samples its typicalness is taken into consideration. It is proved that, independently of the metrics used, the classification error in case of modified distances is less than or equal to the error in case of the nonmodified distances. Furthermore fuzzy information is used to form fuzzy decision rules. That is done to increase the classification accuracy and to establish reasonable limits for fuzzy approach applications. The proposed version is tested with clinical data of neonates, which suffer from Respiratory Distress Syndrome (RDS). On the basis of theoretical and experimental results a computer consulting system for RDS diagnostics is designed.

Cluster ensembles are deemed to be a robust and accurate alternative to single clustering runs. 2... more Cluster ensembles are deemed to be a robust and accurate alternative to single clustering runs. 24 methods for designing cluster ensembles are compared here using 24 data sets, both artificial and real. Adjusted Rand index and classification accuracy are used as accuracy criteria with respect to a known partition assumed to be the "true" one. The data sets are randomly chosen to represent medium-size problems arising within a variety of biomedical domains. Ensemble size of 10 was considered. It was found that there is a significant difference among the compared methods (Friedman's Two Way ANOVA). The best ensembles were based on k-means individual clusterers. Consensus functions interpreting the consensus matrix of the ensemble as data, rather than similarity, were found to be significantly better than the traditional alternatives, including CSPA and HGPA.

Information Fusion, Sep 1, 2006

Adjusted Rand index is used to measure diversity in cluster ensembles and a diversity measure is ... more Adjusted Rand index is used to measure diversity in cluster ensembles and a diversity measure is subsequently proposed. Although the measure was found to be related to the quality of the ensemble, this relationship appeared to be non-monotonic. In some cases, ensembles which exhibited a moderate level of diversity gave a more accurate clustering. Based on this, a procedure for building a cluster ensemble of a chosen type is proposed (assuming that an ensemble relies on one or more random parameters): generate a small random population of cluster ensembles, calculate the diversity of each ensemble and select the ensemble corresponding to the median diversity. We demonstrate the advantages of both our measure and procedure on 5 data sets and carry out statistical comparisons involving two diversity measures for cluster ensembles from the recent literature. An experiment with 9 data sets was also carried out to examine how the diversity-based selection procedure fares on ensembles of various sizes. For these experiments the classification accuracy was used as the performance criterion. The results suggest that selection by median diversity is no worse and in some cases is better than building and holding on to one ensemble.