Milan Tuba

Enhanced firefly algorithm for constrained numerical optimization

Firefly algorithm is one of the recent and very promising swarm intelligence metaheuristics for tackling hard optimization problems. While firefly algorithm has been proven on various numerical and engineering optimization problems as a robust metaheuristic, it was not properly tested on a wide set of constrained benchmark functions. We performed testing of the original firefly algorithm on a set of standard 13 benchmark functions for constrained problems and it exhibited certain deficiencies, primarily insufficient exploration during early stage of the search. In this paper we propose enhanced firefly algorithm where main improvements are correlated to the hybridization with the exploration mechanism from another swarm intelligence algorithm, introduction of new exploitation mechanism and parameter-based tuning of the exploration-exploitation balance. We tested our approach on the same standard benchmark functions and showed that it not only overcame weaknesses of the original firefly algorithm, but also outperformed other state-of-the-art swarm intelligence algorithms.

Drone Placement for Optimal Coverage by Brain Storm Optimization Algorithm

Unmanned aerial vehicles or drones are used in wide range of applications and one of them is area monitoring. Finding the optimal positions for drones so that the coverage is maximized, while reducing the fuel consumption represents computationally hard problem. For these kinds of problems, swarm intelligence algorithms have been successfully used. In this paper we propose recent brain storm optimization algorithm for finding the locations for static drones. Optimal drone placement maximizes the number of covered targets while minimizing drones altitude. The proposed method was tested in two different environments, with uniformly and clustered deployed targets. Based on the obtained results it can be concluded that brain storm optimization is appropriate for solving drone placement problem in both considered environments.

Support Vector Machine Optimized by Elephant Herding Algorithm for Erythemato-Squamous Diseases Detection.

Abstract Machine learning algorithms are used in numerous field and medicine is one of them. Automatic diagnosis or detection of different diseases based on list of symptoms can drastically improve and speedup diagnostics process. Determining diagnosis at earlier stages gives better healing results. In this paper a method for automatic erythemato-squamous diseases classification was proposed. Six erythemato-squamous diseases that are very hard to distinguish were classified by the optimized support vector machine. Recent swarm intelligent algorithm, elephant herding optimization algorithm was used to find optimal parameters for the support vector machine that was then used to determine the exact erythemato-squamous diseases. We compared accuracy of our proposed method to other approaches from literature using standard dataset and it obtained better results in all experiments.

Digital image forgery detection based on shadow HSV inconsistency

During the last decade digital images have spread to all facets of human life. One of the main advantages of digital images is wide availability of powerful digital image processing tools. However, this power and availability also facilitates simplicity of forgery of digital images where it is very easy to insert part of one image into another image. Digital image forensics has to deal with such situations. In this paper we propose an algorithm for digital image forgery detection based on shadow inconsistencies of HSV components. The advantage of using these features is rotational invariance and simplicity. We tested our proposed algorithm on forged images used in literature and the algorithm was successful in forgery detection in all cases.

Web Intelligence Data Clustering by Bare Bone Fireworks Algorithm Combined with K-Means

Data mining and clustering are important elements of various applications in different fields. One of the areas were clustering is rather frequently used is web intelligence, which nowadays represents an important research area. Data collected from the web are usually very complex, dynamic, without structure and rather large. Traditional clustering techniques are not efficient enough and need to be improved. In this paper, we propose combination of recent swarm intelligence algorithm, bare bones fireworks algorithm, and k-means for clustering web intelligence data. The proposed method was compared with other approaches from literature. Based on the experimental results, it can be concluded that the proposed method has very promising characteristics in terms of the quality of clustering, as well as the execution time.

Bare Bones Fireworks Algorithm for Capacitated p-Median Problem.

The p-median problem represents a widely applicable problem in different fields such as operational research and supply chain management. Numerous versions of the p-median problem are defined in literature and it has been shown that it belongs to the class of NP-hard problems. In this paper a recent swarm intelligence algorithm, the bare bones fireworks algorithm, which is the latest version of the fireworks algorithm is proposed for solving capacitated p-median problem. The proposed method is tested on benchmark datasets with different values for p. Performance of the proposed method was compared to other methods from literature and it exhibited competitive results with possibility for further improvements.

Elephant Herding Optimization Algorithm for Wireless Sensor Network Localization Problem

This paper presents elephant herding optimization algorithm (EHO) adopted for solving localization problems in wireless sensor networks. EHO is a relatively new swarm intelligence metaheuristic that obtains promising results when dealing with NP hard problems. Node localization problem in wireless sensor networks, that belongs to the group of NP hard optimization, represents one of the most significant challenges in this domain. The goal of node localization is to set geographical co-ordinates for each sensor node with unknown position that is randomly deployed in the monitoring area. Node localization is required to report the origin of events, assist group querying of sensors, routing and network coverage. The implementation of the EHO algorithm for node localization problem was not found in the literature. In the experimental section of this paper, we show comparative analysis with other state-of-the-art algorithms tested on the same problem instance.

Color Image Segmentation by Multilevel Thresholding Based on Harmony Search Algorithm

One of the important problems and active research topics in digital image precessing is image segmentation where thresholding is a simple and effective technique for this task. Multilevel thresholding is computationally complex task so different metaheuristics have been used to solve it. In this paper we propose harmony search algorithm for finding optimal threshold values in color images by Otsu’s method. We tested our proposed algorithm on six standard benchmark images and compared the results with other approach from literature. Our proposed method outperformed other approach considering all performance metrics.

Hybridized Elephant Herding Optimization Algorithm for Constrained Optimization

This paper introduces hybridized elephant herding optimization algorithm (EHO) adopted for solving constrained optimization problems. EHO is one of the latest swarm intelligence metaheuristic and the implementation of the EHO for constrained optimization was not found in literature. In order to evaluate the performance of the hybridized EHO algorithm, we conducted tests on 13 standard constrained benchmark functions. To prove efficiency and robustness of the hybridized EHO, a comparative analysis with basic EHO implementation, as well as with other state-of-the-art algorithms, such as firefly algorithm, seeker optimization algorithm and self-adaptive penalty function genetic algorithm was performed. Experiments show that the hybridized EHO on average outperforms other algorithms used in comparative analysis.