Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Radoslava Kraleva has a total of 32 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Analyzing the Effect of Recursive Branch Pruning and Heuristic Upper-Bound Initialization in Exact Travelling Salesman Problem Search

    by Velin Kralev, Radoslava Kraleva

    This study investigates exact algorithms for solving the Travelling Salesman Problem (TSP), formulated as the problem of finding a minimum Hamiltonian cycle in a complete weighted undirected graph. Two recursive algorithms are presented and analyzed. The first algorithm, denoted RecursiveHamiltonianCycle (RHC), performs an exhaustive recursive search of the solution space, while the second algorithm, denoted PrunedHamiltonianCycleSearch (PHCS), applies an early branch-pruning mechanism to eliminate non-promising partial solutions. The main objective of this research is to evaluate the influence of recursive branch pruning and initial upper-bound selection on the computational efficiency of exact TSP search procedures. A series of computational experiments was conducted using complete graphs containing between 12 and 21 vertices. The obtained results show that both algorithms always produce identical optimal Hamiltonian cycles. However, the PHCS algorithm requires significantly fewer recursive calls and substantially less execution time. Additional experiments demonstrate that the quality of the initial upper bound has a substantial impact on pruning efficiency. A nearest-neighbor heuristic was therefore used to generate high-quality initial solutions, resulting in further reductions in the search space. The results confirm that combining heuristic initialization with exact branch-pruning search significantly improves the practical applicability of exact TSP algorithms while preserving solution optimality.

  • Open Access

    Improving Exact TSP Search by Recursive Branch Pruning and Heuristic Upper-Bound Initialization

    by Velin Kralev, Radoslava Kraleva

    This study investigates exact algorithms for solving the Travelling Salesman Problem (TSP), formulated as the problem of finding a minimum Hamiltonian cycle in a complete weighted undirected graph. Two recursive algorithms are presented and analyzed. The first algorithm, denoted RecursiveHamiltonianCycle (RHC), performs an exhaustive recursive search of the solution space, while the second algorithm, denoted PrunedHamiltonianCycleSearch (PHCS), applies an early branch-pruning mechanism to eliminate non-promising partial solutions. The main objective of this research is to evaluate the influence of recursive branch pruning and initial upper-bound selection on the computational efficiency of exact TSP search procedures. A series of computational experiments was conducted using complete graphs containing between 12 and 21 vertices. The obtained results show that both algorithms always produce identical optimal Hamiltonian cycles. However, the PHCS algorithm requires significantly fewer recursive calls and substantially less execution time. Additional experiments demonstrate that the quality of the initial upper bound has a substantial impact on pruning efficiency. A nearest-neighbor heuristic was therefore used to generate high-quality initial solutions, resulting in further reductions in the search space. The results confirm that combining heuristic initialization with exact branch-pruning search significantly improves the practical applicability of exact TSP algorithms while preserving solution optimality.

  • Open Access

    A systematic review of the design and artefacts of contemporary illustrated children's books in Bulgaria

    by Radoslava Kraleva, Victoria Kraleva

    This article examines the characteristics of contemporary illustrated children’s books published in Bulgarian, approaching them as literary, visual, and publishing artefacts. The analysis is based on a corpus of 182 titles released between 2022 and 2025 and is not intended as a comprehensive representation of the national book market. The study constitutes a systematic review that integrates quantitative and qualitative examination of authorship, the material properties of the book object, market prices, text organization, interactive features, accessibility for children with special educational needs, and the main character types. The findings reveal a predominance of hardback editions, an average length of approximately 32 pages, and a gradual increase in the mean market price. Although foreign authors prevail within the sample, a steady growth in the number of Bulgarian writers is evident, alongside an increasing tendency for authors to illustrate their own books. The analysis identifies a preference for anthropomorphic animals, fantastical characters, and formats that support visual perception, shared reading, early literacy, and empathy. A growing use of accessibility-oriented elements, such as specialized typefaces and audio versions of the text, is also observed. Overall, the study demonstrates that contemporary children’s books in Bulgaria operate as a complex cultural and market product in which content, visual design, material production, and price positioning are closely interrelated.

  • Open Access

    Evaluating the Influence of Graph Density on the Efficiency of Shortest Path Algorithms Using Different Data Structures

    by Velin Kralev, Radoslava Kraleva, Aleksandra Popova

    This paper presents a comparative analysis of two variants of a classical algorithm for finding the shortest path in a connected graph. The first variant uses an adjacency matrix (AM) to verify the existence of an edge (arc) between two vertices, while the second variant performs the same verification using an adjacency list (AL). The objective of this study is to examine how graph density affects the performance of the two algorithmic modifications depending on the data structure used. A total of 95 graphs were analyzed, grouped into five sets from 100 to 500 in increments of 100. For each group, 19 graphs were generated with densities ranging from 5% to 95% in increments of 5%. The methodology includes analyzing the number of iterations, assignments, and comparisons executed by the algorithms for all graphs. The initial hypothesis assumed that the total number of operations would always be lower when using an AL instead of an adjacency matrix, regardless of graph density. The results demonstrate that this assumption is incorrect: for densities above 82%, the total number of operations is lower when using an adjacency matrix, whereas the AL is more efficient for densities below 82%, with its efficiency increasing as density decreases. These findings are particularly important for mobile technologies, as they support the design of more efficient pathfinding solutions that optimize performance and energy consumption in mobile applications.

  • Open Access

    Influence of the graph density on approximate algorithms for the graph vertex coloring problem

    by Velin Kralev, Radoslava Kraleva

    This research explores two heuristic algorithms designed to efficiently solve the graph coloring problem. The implementation codes for both algorithms are provided for better understanding and practical application. The experimental methodology is thoroughly discussed to ensure clarity and reproducibility. The execution times of the algorithms were measured by running the test applications six times for each analyzed graph. The results indicate that the first algorithm generally produced better solutions than the second. In only two instances did the first algorithm produce solutions comparable to those of the second. The results reveal another trend: as the graph density exceeds 85%, the number of required colors increases significantly for both algorithms. However, even at a density of 95%, the number of colors required to color the graph's vertices does not exceed half the total number of vertices. As the graph density increases from 95% to 100%, the number of colors required to color the graph rises significantly. However, when the graph density exceeds 97%, both algorithms produce identical solutions.

  • Open Access

    An analysis between the Welsh-Powell and DSatur algorithms for coloring of sparse graphs

    by Radoslava Kraleva, Velin Kralev, Toma Katsarski

    In this research an analysis between the Welsh-Powell and DSatur algorithms for the graph vertex coloring problem was presented. Both algorithms were implemented and analyzed as well. The method of the experiment was discussed and the 46 test graphs, which were divided into two sets, were presented. The results show that for sparse graphs with a smaller number of vertices and edges, both algorithms can be used for solving the problem. The results show that in 50% of the cases the Welsh-Powell algorithm found better solutions (23 in total). So, the DSatur algorithm found better solutions in only 19.6% of cases (9 in total). In the remaining 30.4% of cases, both algorithms found identical solutions. For graphs with a larger number of vertices, the usage of the Welsh-Powell algorithm is recommended as it finds better solutions. The execution time of the DSatur algorithm is greater than the execution time of the Welsh-Powell algorithm, reaching up to a minute for graphs with a larger number of vertices. For graphs with fewer vertices and edges, the execution times of both algorithms are shorter, but the time is still greater for the DSatur algorithm.

  • Open Access

    A Web Service-Based Application for Processing EEG Experimental Data Generated in Response to Visual Stimuli

    by Velin Kralev, Radoslava Kraleva, Petia Koprinkova-Hristova, Nadejda Bocheva, Miroslava Stefanova

    This paper presents the results of a comparative analysis of different approaches to retrieving experimental data. The different methods of processing the data and their applications are analyzed. For the aims of the study, a database was designed to store, analyze, and process experimental data generated in response to visual stimuli. The methodology and conditions for conducting the experiments are detailed. The experiments utilized real data, which can be extracted by calling the corresponding web methods. For the experiments, 12 independent tests were conducted, and six operations related to extracting, processing, and analyzing the experimental data were examined. The results indicate that data extraction is fastest when using a direct approach, specifically by executing a query directly on the database server. Based on the results, the two slowest operations were found to be related to data extraction using web methods. The first variant describes the data in text format, while the second uses XML. The second approach is significantly slower in data retrieval, by about 30%. In addition to data retrieval operations, another data processing operation was also analyzed once the data was buffered in the address space of the client application. It was found that these operations are performed in an acceptable time—slower than direct data retrieval from the database server but faster than data retrieval using web methods.

  • Open Access

    Combining Genetic Algorithm with Local Search Method in Solving Optimization Problems

    by Velin Kralev, Radoslava Kraleva

    This research is focused on evolutionary algorithms, with genetic and memetic algorithms discussed in more detail. A graph theory problem related to finding a minimal Hamiltonian cycle in a complete undirected graph (Travelling Salesman Problem—TSP) is considered. The implementations of two approximate algorithms for solving this problem, genetic and memetic, are presented. The main objective of this study is to determine the influence of the local search method versus the influence of the genetic crossover operator on the quality of the solutions generated by the memetic algorithm for the same input data. The results show that when the number of possible Hamiltonian cycles in a graph is increased, the memetic algorithm finds better solutions. The execution time of both algorithms is comparable. Also, the number of solutions that mutated during the execution of the genetic algorithm exceeds 50% of the total number of all solutions generated by the crossover operator. In the memetic algorithm, the number of solutions that mutate does not exceed 10% of the total number of all solutions generated by the crossover operator, summed with those of the local search method.


  • Open Access

    An analysis between different algorithms for the graph vertex coloring problem

    by Velin Kralev, Radoslava Kraleva

    This research focuses on an analysis of different algorithms for the graph vertex coloring problem. Some approaches to solving the problem are discussed. Moreover, some studies for the problem and several methods for its solution are analyzed as well. An exact algorithm (using the backtracking method) is presented. The complexity analysis of the algorithm is discussed. Determining the average execution time of the exact algorithm is consistent with the multitasking mode of the operating system. This algorithm generates optimal solutions for all studied graphs. In addition, two heuristic algorithms for solving the graph vertex coloring problem are used as well. The results show that the exact algorithm can be used to solve the graph vertex coloring problem for small graphs with 30-35 vertices. For half of the graphs, all three algorithms have found the optimal solutions. The suboptimal solutions generated by the approximate algorithms are identical in terms of the number of colors needed to color the corresponding graphs. The results show that the linear increase in the number of vertices and edges of the analyzed graphs causes a linear increase in the number of colors needed to color these graphs.


  • Open Access

    A comparative analysis between two heuristic algorithms for the graph vertex coloring problem

    by Velin Kralev, Radoslava Kraleva

    This study focuses on two heuristic algorithms for the graph vertex coloring problem: the sequential (greedy) coloring algorithm (SCA) and the Welsh–Powell algorithm (WPA). The code of the algorithms is presented and discussed. The methodology and conditions of the experiments are presented. The execution time of the algorithms was calculated as the average of four different starts of the algorithms for all analyzed graphs, taking into consideration the multitasking mode of the operating system. In the graphs with less than 600 vertices, in 90% of cases, both algorithms generated the same solutions. In only 10% of cases, the WPA algorithm generates better solutions. However, in the graphs with more than 1,000 vertices, in 35% of cases, the WPA algorithm generates better solutions. The results show that the difference in the execution time of the algorithms for all graphs is acceptable, but the quality of the solutions generated by the WPA algorithm in more than 20% of cases is better compared to the SC algorithm. The results also show that the quality of the solutions is not related to the number of iterations performed by the algorithms.