Search

This paper proposes a novel edge-detection, edge preservation and smoothening filter for infrared images. The proposed method consists of two steps: extracting the edges of Infrared (IR) image using modified canny operator and edge-preserving and smoothening using bilateral/guided filter. Experimental result shows the effectiveness of the proposed method in terms of the visual quality.

Through this paper performance of the basic education system of Myanmar is assessed, laying emphasis on secondary education concerning science and mathematics. The assessment tool used is Data Envelopment Analysis (DEA), a linear programming-based method, a useful tool for evaluating the relative efficiency of decision-making units (DMUs) through different input and output parameters.

As compared to general RGB images which contain only three bands from visible range of electromagnetic spectra, hyperspectral images contain several spectral bands covering a larger area of electromagnetic spectra including the infrared and UV. Hence, the hyperspectral data is normally huge and high dimensional. Hyperspectral imaging is an emerging area of research having applications in diverse domains, and through this paper, the authors try to bring forward some recent developments in the area of hyperspectral imaging while using metaheuristics. The authors have tried to identify the various domains where the metaheuristics-embedded hyperspectral imaging has been successfully implemented.

Over the last few years, meta-heuristics algorithms are very popular for solving different types of optimization problems in various fields. In the present study, three nature-inspired algorithms (PSO, ABC, CS), two evolutionary algorithms (GA, DE) and one classical method (SA) are applied in branched network problem to find optimal design cost of network. The branched network consists of two types of problems: first is single load branched network problem and second is multiple load branched network problem.

Elementary education is crucial for the country’s economic growth. It is an instrument we use as a mentor to be successful in life. In this paper, we study the school-level data for various states in India which includes union territories also and collected from government UDISE statistics. In this paper, we have evaluated efficiency using data envelopment analysis (DEA) and multi-objective differential evolution (MODE) with adaptive constraint optimization methods. The data set used in the study includes the basic needs of schools and enrolment of students. Based on the certain input and output parameters, our analysis shows that in how many states the education delivery system is accurate and utilizing the resources provided by the government.

Fuzzy C-Means (FCM) is the most commonly used clustering algorithm despite being sensitive to initial choice of the cluster centroids. To overcome this shortcoming, the present study proposes a clustering algorithm called Fuzzy Particle Swarm Page Rank Clustering (Fuzzy-PSOPRC) which embeds the concept of fuzzy clustering into particle swarm optimization. Fuzzy-PSOPRC gives more precise cluster centroids which groups data objects into better clusters. The proposed algorithm has been tested and evaluated on benchmark datasets. Experimental results show that the proposed clustering algorithm achieves better results in comparison with other clustering techniques in terms of accuracy and F1 measure.

Most of the real-world search and optimization problems have multiple objectives. In the past few years, Multi-objective Evolutionary Algorithms (MOEAs) are widely used. A multi-objective optimization problem has several conflicting objectives that provide a set of trade-off optimal solutions. In this paper, we provide a brief review of Multi-objective Differential Evolution (MODE) primarily during the last 5 years. It covers several improved and modified variants of DE for solving multi-objective engineering problems.

The Indian Paper Industry (IPI) uses diverse fibers raw materials for pulp and papermaking process. The consumption of these fibers raw material depends upon the availability of the raw material, grade of the paper to be produced, and the environmental factors. However, the quality of the paper produced entirely depends upon the quality of the pulp obtained through selected fibers. The pulp quality of the fiber can be determined by the morphological properties of that fiber. Thus, in this paper, the morphological properties of the fiber are prioritized by the customer preference rating (CPR) method, and then further selection of raw materials is performed using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS). A case study of some woods-based fiber raw materials is also presented to prove the applicability and efficacy of the proposed methodology.

Optimization problems can be articulated by numerous practical problems. These glitches stance a test for the academics in the proposal of proficient procedures skilled in digging out the preeminent elucidation with the slightest computing cost. In this study, we worked on differential evolution and cultural algorithm, conglomerates the features of both the algorithms, and proposes a new evolutionary algorithm. This jointure monitors the complex collaboration amalgam of two evolutionary algorithms, where both are carried out in analogous. The novel procedure termed as CA-DE accomplishes an inclusive inhabitant that is pooled among both metaheuristics algorithms concurrently. The aspect of the recycled approval action in credence space is to update the information of the finest individuals with the present information. This collective collaboration arises among both the algorithms and is presented to mend the eminence of resolutions, ahead of the individual performance of both the algorithms. We have applied the newly proposed algorithm on a set of six standard benchmark optimization problems to evaluate the performance. The comparative results presented demonstrate that CA-DE has an encouraging accomplishment and expandable conducts while equated with new contemporary advanced algorithms.

The present study proposes TLBO-PSO an integrated Teacher–Learning-Based Optimization (TLBO) and Particle Swarm Optimization (PSO) for optimum data clustering. TLBO-PSO algorithm searches through arbitrary datasets for appropriate cluster centroid and tries to find the global optima efficiently. The proposed TLBO-PSO is analyzed on a set of six benchmark datasets available at UCI machine learning repository. Experimental result shows that the proposed algorithm performs better than the other state-of-the-art clustering algorithms.

The media is known as a mirror of any society, and because of that, a national newspaper can be analyzed to know the current mindset of any nation. The different types of news and reports actually show the countrymen’s interests as media cover majority of that news that is of interest to their readers. In this study, the analysis is performed on the two leading Indian news providers, known as Hindustan Times and India Today. The main aim of this study is to find out the interest of Indians and Indian media houses in terms of national and international news. To make a generalized comment, six-month data of recent past is used for the analysis purpose. Different parameters (for same duration) are considered for the analysis of the news for both the e-papers so that media house’s interest can also be compared.

Evolutionary computation has turned into a significant problem-solving approach among several researchers. As compared to other existing techniques of global optimization, the population-based combined learning procedure, robustness, and self-adaptation are some of the vital topographies of evolutionary algorithms. In spite of evolutionary algorithms has been broadly acknowledged for resolving numerous significant real applications in various areas; however in practice, occasionally they carry only fringe performance. There is slight motivation to assume that one can discover an unvaryingly finest optimization algorithm for resolving all optimization problems. Evolutionary algorithm depiction is resolute by the manipulation and survey liaison retained during the course. All this evidently elucidates the necessity for fusion of evolutionary methodologies, and the aim is to enhance the performance of direct evolutionary approach. Fusion of evolutionary algorithms in recent times is gaining popularity owing to their proficiencies to resolve numerous legitimate problems such as, boisterous environment, fuzziness, vagueness, complexity, and uncertainty. In this paper, first we highlight the necessity for fusion of evolutionary algorithms and then we explain the several potentials of an evolutionary algorithm hybridization and also discuss the general architecture of evolutionary algorithm’s fusion that has progressed all through the recent years.

Portfolio optimization plays an important role in managing the financial assets of an individual. The investments are made such that an individual attains the maximum benefit out of it. In this paper, a bi-objective portfolio optimization model is considered, where the objectives are to maximize the return and minimize the risk, and is solved using multi-objective synergetic differential evolution (MO-SDE).

Accurate and unambiguous measurement of relative displacement between two images is important in a large number of practical applications. Comparison of a real-time captured image with a stored reference image is one such example. Comparison of two sensed images acquired within a short and accurately known time interval can be used to measure the velocity to height ratio of a moving imaging system. Accurate superposition of two images acquired at different times and their consequent, point by point subtraction, calls for attention to changes which have occurred over time in the real-time scenario. All such applications require measurement of the displacement vector to accuracies within a small fraction of pixel. The phase correlation method is based on the fact that most of the information about the relative displacement vector is contained in the phase of their cross-power spectrum.

A lot of science and engineering problems are nonlinear in nature. Due to various limitations of the traditional methods, we need some robust and efficient non-traditional techniques to solve these nonlinear and complex problems. In the current study, a new variant of differential evolution (DE) algorithm named information utilization-based modified differential evolution (IUMDE) algorithm is proposed. Further the proposed variant is implemented on six benchmark problems which are taken from literature. Experiments, results, and comparison confirm the competence and effectiveness of the proposed variant algorithm over the others.

For promoting products in a highly competitive market, it is imperative for firms to develop appropriate promotions. Advertising is one of the major components of promotional mix which contributes in communicating the distinct product features to customers. Multiple media, namely television, radio, mobile phones, newspaper, magazines, Web sites, are available to the firms for advertising. Although over the years there have been many new media that have been included in the media mix of firms, yet television has remained important to advertisers. This importance is owing to its features such as large reach to the audience as well as effective communication with the audience. This paper presents a review of literature in the field of scheduling of advertisements in television.

Past one decade has seen a tremendous increase in the development of nature-inspired metaheuristics (NIM). In this paper, we provide a brief review on firefly algorithm (FA), one of the recent developments in the field of NIM for solving optimization problems. FA first proposed by Xin-She Yang in 2008 is a swarm intelligence-based algorithm slowly gaining popularity for solving continuous, multi-objective, dynamic and noisy, discrete, nonlinear, and multi-dimensional problem. It has a simple structure and is easy to apply on a wide range of problems as shown by different researchers.

This paper estimates the technical efficiencies of 20 departments of a government PG college Gopeshwar, Chamoli, Uttarakhand (India), for the academic year 2012–2013 through DEA technique. Number of academic staffs, number of taught course students, average students’ qualification, departmental operating cost (DOC) are taken as input variables, and number of graduates from taught courses, average graduates’ results (%), and graduate rate (%) are taken as output variables. Input–output data are collected from the NAAC report 2014 and annual SSR report academic year 2012–2013 and from the examination section of the government PG college, Gopeshwar. Some data are collected from the departments of the college and college annual book Madhuri. The study concludes that overall efficiency of departments is 87.8%, which indicates that the departments are not using their potential by 12.20%.

This paper proposes the comparative analysis of accuracy and efficiency in patch-based texture synthesis among different color models and SSD (sum of squared distance) method. Here we have taken the basic texture quilting method of Efros and Freeman as a synthesizer for analysis. To find the “best-fit” patch for synthesizing the image, this paper shows these different color models and SSD method on randomly fetched patch. Finally comparison is done on these different methods to analyze the result and synthesized images have shown for better visualization.

Shuffled frog-leaping algorithm (SFLA), a memetic algorithm modeled on the foraging behavior of natural species called frogs. SFLA embeds the features of both particle swarm optimization (PSO) and shuffled complex evolution (SCE) algorithm. It is well documented in literature that SFLA is an efficient algorithm to solve non-traditional optimization problems. However like other memetic algorithms SFLA also limits in convergence rate or shows premature convergence when applied to multifaceted continuous optimization problems. In this study, an opposition-based variant of SFLA named as O–SFLA is proposed. In general, the structure of SFLA, the frog, is divided into memeplexes based on their fitness values where they forage for food. In this study, the opposition-based learning concept is embedded into the memeplexes before the frog initiates foraging. The proposed variant is validated on six optimization benchmark problems taken from literature. Further non-parametric analysis is performed to evaluate the efficacy of the proposal.

Shuffled frog-leaping algorithm (SFLA) is comparatively a recent addition to the family of nontraditional population-based search methods that mimics the social and natural behavior of species (frogs). SFLA merges the advantages of particle swarm optimization (PSO) and genetic algorithm (GA). Though SFLA has been successfully applied to solve many benchmark and real-time problems it limits the convergence speed. In order to improve its performance, the frog with the best position in each memeplexes is allowed to slightly modify its position using random walk. This process improves the local search around the best position. The proposal is named improved local search in SFLA (ILS-SFLA). For validation, three engineering optimization problems are consulted from the literature. The simulated results defend the efficacy of the proposal when compared to state-of-the-art algorithms.

Automobile industries worldwide are unified in opinion, that successful management of sustainable supply chains is the most important driver to improve both their economic and ecological performances. The significance of sustainable supply chain management (SSCM) is a critical corporate matter in the automobile industries that offers incredible potential for achieving better environmental performance, consumer fulfillment, pull down operating expenditures, reducing inventory investments in addition to achieving better fixed asset usage. The environment concerns, climatic changes, and additional ecological concerns in automobile industries are not only articulated by campaigners or researchers, but also by the common man as well, which has motivated the industries to focus on sustainability. The present research focuses on a DEA-based mathematical model and employs differential evolution to select the competent suppliers providing the utmost fulfillment for the sustainable criteria determined. This study aims to examine the sustainable supplier evaluation and selection practices likely to be adopted by the Indian automobile industry for their products.

Portfolio optimization guides about the management of assets. Among several investment offers the task is to choose the plan so as to attain the maximum financial benefit. The investor requires a thorough comparative study to decide the best possible option where the return is maximum and the risk is minimum. Portfolio optimization can help in the process of decision-making by bringing out the selected options beneficial for the investor. In this paper we have considered the mean semivariance portfolio optimization model given by Markovitz and the data is taken from National Stock Exchange (NSE), Mumbai. Ten years data of ten different banks is taken from national stock exchange. The model is solved with the help of differential evolution which is a population-based metaheuristics.

The process of medical diagnosis, like many other fields, has to pass through various stages of uncertainty, especially in cases where the data is mostly available in linguistic format. Under such circumstances of vague data, application of fuzzy logic concepts can play an important role in extracting approximate information which in turn may help in reaching to a particular diagnosis. This study is devoted to the application of fuzzy logic in the psychological domain. The paper provides a detailed literature review on the use of fuzzy logic rules in analyzing the different aspects of psychological behavior of human beings. Further, it also provides some suggestions to make the system more effective.

In this paper, we have studied nonlinear optical properties like second harmonic generation of organic composite polymeric films. Composite films are prepared using organic host polymer polyethersulfone and organic guest material metanitroaniline. Films are poled by contact poling method to attain the desired noncentrosymmetry. Poled films are characterized by Nd:YAG laser to analyze the second harmonic generation efficiency. Second harmonic intensity of films are evaluated applying ‘fmincon’ statistical tool, as a function of laser input power and metanitroaniline doping concentration.

The present study proposed a robust and secure watermarking scheme to authenticate the digital images for ownership claim. The proposed watermarking scheme is making use of 2-level of DWT (Discrete wavelet transform) to provide high capacity of watermark embedding. The SVD (singular value decomposition) is performed on the host and watermark images. Then, principal components are calculated for watermark image. The use of principal components for watermark embedding makes the scheme free from false positive error. PSO (particle swarm optimization) optimized multiple scaling factors along with principal components are utilized for watermark embedding in the singular values of host image. The PSO-optimized scaling factor provides a very good tradeoff between imperceptibility and robustness of watermarking scheme. The scheme is also extended to use for color images. The proposed scheme provides a secured and high data embedding with good robustness toward different signal processing attacks.