Computational Collective Intelligence

The paper describes the derivation of a recursive P-spline difference equation. It further demonstrates how a transformed spline can be used as a digital filter with an infinite impulse response and variable parameters. Frequency responses of a real-time spline filter correspond to frequency responses of low-frequency digital filters. The paper further examines the influence of some P-spline parameters on the efficiency of interpretation of real-time input measurement information.

In recent literature from Neuroscience the adaptive role of the effects of stress on decision making is highlighted. The problem addressed in this paper is how that can be modelled computationally. The presented adaptive temporal-causal network model addresses the suppression of the existing network connections in a first phase as a result of the acute stress, and then as a second phase relaxing the suppression after some time and give room to start new learning of the decision making in the context of the stress again.

This paper concentrates on the algorithm based on simulated annealing approach. The algorithm was implemented for solving the formulated chess problem. The properties of the algorithm were analyzed taking into account the results of experiments made using the designed and implemented experimentation system. This system allows testing various configurations of the algorithm and comparing the effects with those obtained by the algorithms based on ant colony optimization and genetic evolutionary ideas. The paper shows that the proposed algorithm seems to be promising.

This paper concentrates on the created algorithm for solving the single machine total weighted tardiness problem (SMTWT). The algorithm is based on searching the solution space along with the tree rules. The properties of the algorithm are studied taking into account the results of experiments made using the designed and implemented experimentation system. This system allows testing various configurations of the algorithm as well as comparing the effects obtained by this algorithm with effects of known meta-heuristic algorithms, which are based on Simulated Annealing and Invasive Weed Optimization. The paper shows that the proposed algorithm requires some improvements, however seems to be promising.

This paper proposes The Multi-Layer Strategies Differential Evolution (MLSDE) algorithm, which finds optimal solutions for large scale problems. To solve large scale problems were grouped different strategies together and applied them to date set. Furthermore, these strategies were applied to selected vectors to strengthen the exploration ability of the algorithm. Extensive computational analysis were also carried out to evaluate the performance of the proposed algorithm on a set of well-known CEC 2015 benchmark functions. This benchmark was utilized for the assessment and performance evaluation of the proposed algorithm.

Modern distributed industrial control systems needs to transfer many different types and amounts of information between different devices. One type of information is measurements – analogue information directly from the production level. In order to achieve the technological requirements of a production system, an open standard Open Process Control – Unified Architecture (OPC UA) system was selected. An embedded automation device equipped with an OPC UA server can either transmit deadband tested analogue attributes or all of the changes of the analogue attributes to the OPC UA clients. The subscription interval, sampling interval and deadband are controlled by the OPC UA clients.While the floating point attributes in an automation device change all the time, in real industry systems, the analogue attributes of small changes do not necessarily have to be communicated to the OPC UA clients frequently. In addition to the attribute deadband test, changes in some of the attributes can safely be blocked for a minimum amount of time after the last attribute update before a new value is transmitted to the UA clients.Due to the nature of deadbands in attribute communication, analogue attributes should also be transmitted to the OPC UA clients at a low frequency to make sure that the client receives the correct attribute value after a set maximum time.This paper shows an approach to OPC UA data traffic control using a Programmable Logic Controller (PLC) function block to control traffic of analogue attributes a server and its clients.

The quality management process is one of the most important manufacturing activities. Although it can be implemented in a dedicated IT system, because of the easy access to production data in Manufacturing Execution Systems, it seems that an MES is a particularly convenient place for its implementation. This paper describes the concept of using decision tree methods to analyse the relationship between the production path and quality problems. The proposed method is based on an information model that is compliant with the ISA95 standard. Because of this, it can be applied not only in the case presented in the research part, but is also applicable for the problem of quality analysis in other types of discrete production. The authors present the information model that was used, the proposed method of analysis and the results for the simulation data. The simulation scenario was created as a simplification of the actual production process of electronic devices performed by AIUT company.

This paper present the heuristic algorithm that facilitates very efficient creation of personalised timetables for university students. The goal of the algorithm is to find a timetable with the minimal number of event overlaps and at the same time optimizing the length of the stay of the student at university every week. This goal has been achieved by the modular design of the algorithm that gradually constructs the timetable together with pruning the space of remaining options. The results delivered by this algorithm were compared with the results obtained by exhaustive search and it was concluded that our algorithm is able to find the optimal or a very near optimal solution very quickly. Therefore, the algorithm described here can be utilized for the development of the mobile application.

The paper presents an Evolutionary Algorithm (EA) based framework capable of handling a variety of complex Multi-Robot Task Allocation (MRTA) problems. Equipped with a flexible chromosome structure, customized variation operators, and a penalty function, the EA demonstrates the capability to switch between single-robot and multi-robot cases of MRTA and entertains team heterogeneity. The framework is validated and compared against a Genetic Algorithm based representation and a heuristic-based solution. The experimental results show that the presented EA provides better overall results to the task allocation problem with faster convergence and lesser chances of sub-optimal results.

A prediction mechanism for Memetic Algorithm is presented in this paper. The Predictive Memetic Algorithm (PMA) uses a nonlinear regression method to estimate the parameters used by the algorithm to obtain good solutions in a dynamic and stochastic environment. The algorithm is applied to nonlinear data sets and performance is compared with genetic and simulated annealing algorithms. When compared with the existing methods, the proposed method generates a relatively small error difference after prediction thereby proving its superior performance. A dynamic stochastic environment is used for experimentation, so as to determine the efficacy of the algorithm on non-stationary problem environments.

We propose a novel strategy of decision-making based on the idea of the position in which different players find themselves in a board game to focus not only on own piece but also on all pieces on the same position. This strategy will be independent of any particular search algorithm, thereby providing good quality movement for a general-purpose player. In an attempt to provide more insight into the nature of modeling artificial players three algorithms and five strategies in total have been implemented in the Great Turtle Race game. Based on statistical analysis the highest winning rate is found using this positioning strategy combined with alpha-beta pruning. In particular, this paper presents the joint model of these algorithms and strategies together with a concise summary of the game rules, suggesting possible correlations. These theoretical findings are complemented by experiments that were conducted to evaluate the winning rates.

This article deals with the problem of partitioning a space between a number of robots in a distributed and dynamic way. We aim through the proposed approach to divide the area of interest into a number of sub-regions of equal sizes using the Voronoi diagram. There is no central control, and the robots operate in a completely autonomous way, from the neighborhood discovery to the localization and position sharing. The individual actions of the robots are controlled by the Belief Desire Intention (BDI) model which allows them to operate deliberately and readjust their plans on the go, making the system evolve dynamically.We show in this paper, through a series of conducted experiments, the results of the proposed approach for different maps with different number of robots and the advantage of the use of the BDI model that makes the robots instantly readjust their calculations when a change occurs on the network.

Swarm intelligence offers useful instruments for developing collective behaviours to solve complex, ill-structured and large-scale problems. Efficiency in collective behaviours depends on how to harmonise the individual contributions so that a complementary collective effort can be achieved to offer a useful solution. The harmonisation helps blend diversification and intensification suitably towards efficient collective behaviours. In this study, two renown honeybees-inspired algorithms were analysed with respect to the balance of diversification and intensification and a hybrid algorithm is proposed to improve the efficiency accordingly. The proposed hybrid algorithm was tested with solving well-known highly dimensional numerical optimisation (benchmark) problems. Consequently, the proposed hybrid algorithm has demonstrated outperforming the two original bee algorithms in solving hard numerical optimisation benchmarks.

This paper presents a fine-grained parallel genetic algorithm improved with a 2-opt heuristic for finding solutions near to the optimum to the Quadratic Assignment Problem (QAP). The proposed algorithm is fully implemented on Graphics Processing Units (GPUs). Unlike previous approaches reported in the literature our implementation a two-dimensional GPU grid of size $$10\,\times \,10$$10×10 defines the population of the genetic algorithm (set of permutations of the QAP) and each GPU block consists of n GPU threads where n is the size of QAP. Each GPU block is used to represent the chromosome of a single individual and each GPU thread represents a gene of such chromosome. The proposed algorithm is tested on a subset of the standard QAPLIB data set. Our results show that our implementation is able to find good solutions for large QAP instances in few parallel iterations of the evolutionary process.

Consensus methods are used mainly to solve conflicts of knowledge in decision support systems. Generally speaking, conflicts of knowledge arise from the fact that system nodes (for example, agents, experts) may present various decisions or solutions to the user. This may be due to the use of various methods of decision support or different information sources by agents/experts. If there is a conflict of knowledge in the system and they are not automatically resolving the system cannot generate the final decision, and hence - the decision maker will not receive hints from the system. The use of consensus methods eliminates this problem, because they enable to solve conflicts of knowledge in near real time. At the same time they guarantee the achievement of a good compromise. However, the effective determination of consensus depends, among other, on the correct definition of the distance function.The aim of this paper is to develop a new distance function between the decisions generated by expert of agents in decision support systems.

In this study, a new optimization algorithm called the Hitchcock Birds Inspired Algorithm (HBIA) is introduced, inspired by the aggressive bird behavior portrayed by Alfred Hitchcock in the 1963 thriller “The Birds”. When gathering elements about the phenomenon of birds throughout the film, it is possible to enumerate characteristics of the behavior of the birds that Hitchcock portrayed in the film. HBIA is a stochastic swarm intelligence algorithm that captures the essence of the fictional behavior of birds exposed by Hitchcock and model an optimization mechanism. The algorithm was based on the attack pattern of birds in the film, which has the stages of stalking, attack and reorganization, defined by the initialization, movement strategies in the search space and strategy of local minimum escape, respectively. The technique has as differential the use of adaptive parameters, a discretized random initialization and the use of the Beta distribution. When comparing to SCA, WOA, TLBO and VS, HBIA’s performance is investigated by several experiments implemented in eight cost functions. The results show that the HBIA can find more satisfactory solutions in high dimensionality in the majority of the evaluated cost functions compared to the other four methods.

The paper focuses on Dynamic Vehicle Routing Problem with Time Windows, which generalizes its static counterpart by assuming that information about customers is not given a priori to the decision maker and it may change during the execution of the routes. Multi-Agent System to simulate and solve DVRPTW proposed by the author in his previous work has been extended in the paper. Taking into account different roles of the agents in the proposed system, two forms of cooperation (vertical and horizontal) between them have been implemented in the system. Whereas vertical cooperation refers to cooperation between different groups of agents, horizontal cooperation focuses on cooperation between agents belonging to the same group and/or working at the same level of the multi-agent system. Positive impact of different forms of cooperation on the results has been confirmed by a computational experiment.

Instance selection, often referred to as data reduction, aims at deciding which instances from the training set should be retained for further use during the learning process. Instance selection is the important preprocessing step for many machine leaning tools, especially when the huge data sets are considered. Class imbalance arises, when the number of examples belonging to one class is much greater than the number of examples belonging to another. The paper proposes a cluster-based instance selection approach for the imbalanced data classification. The proposed approach bases on the similarity coefficient between training data instances, calculated for each considered data class independently. Similar instances are grouped into clusters. Next, the instance selection is carried out. The process of instance selection is controlled and carried-out by the team of agents. The proposed approach is validated experimentally. Advantages and main features of the approach are discussed considering results of the computational experiment.

This paper presents a proof-of concept study for demonstrating the viability of building collaboration among multiple agents through standard Q learning algorithm embedded in particle swarm optimisation. Collaboration is formulated to be achieved among the agents via competition, where the agents are expected to balance their action in such a way that none of them drifts away of the team and none intervene any fellow neighbours territory, either. Particles are devised with Q learning for self training to learn how to act as members of a swarm and how to produce collaborative/collective behaviours. The produced experimental results are supportive to the proposed idea suggesting that a substantive collaboration can be build via proposed learning algorithm.

This paper discusses the so-called multigraph, in which multiple connections (arcs) between two given nodes and loops are possible. The authors present the concept of using Shapley values to analyse both elements (nodes and arcs) of a multigraph. The results obtained allow to evaluate the importance of a given element (node or arc) as an element of a whole structure of the graph. The authors proposed a new cooperative game solution and Shapley value which may determine the evaluation of multigraph elements (which may, among others, determine the cost of its use or the volume of flows).

Integrating affect in both individual and collective decision-making processes in order to solve real-world problems can be challenging. This research aims to: (1) investigate how group affect (moods, emotions, and feelings) can be integrated and formalized in the decision-making processes; (2) develop current practices; and (3) draw ideas for future perspectives and real-world applications. For this purpose, the role of affect in decision-making is investigated on the individual behavior level, emotional intelligence, and the collective behavior level. The used methodology consists of exploring and investigating the main characteristics developed in group affect in complex decision-making systems from psychology to computer science. From this, a common global structure is deduced: individual processes, group processes and emerging processes (bottom-up, top-down, and combination of bottom-up and top-down components). Following this, one psychological model and two computational models of group emotion and decision are analyzed, and discussed. Their different approaches to developing the main characteristics of a computational model integrating group affect in the decision-making process are highlighted. Finally, specific scenarios of real-world applications are presented in order to draw interesting and promising computational model perspectives.

The paper describes a GEP-based ensemble classifier constructed using the stacked generalization concept. The classifier has been implemented with a view to enable parallel processing, with the use of Spark and SWIM - an open source genetic programming library. The classifier has been validated in computational experiments carried-out on benchmark datasets.

In this paper a multi-agent system based on the A-Team concept is proposed to solve the distributed resource-constrained multi-project scheduling problem (DRCMPSP). The DRCMPSP belongs to the class of the strongly NP-hard optimisation problems. In the DRCMPSP multiple distributed projects are considered, hence, a coordination of the shared decisions is needed as well as the local task schedule for each project. Multi-agent systems are the natural way of solving such problems. The proposed A-Team multi-agent system has been built using the JABAT environment where two types of the optimisation agents are involved: local and global. Local agents are used to find solutions for the local projects, and global agents are responsible for coordination of the local projects and hence, for the global solution. The approach has been tested experimentally using 140 benchmark problem instances from MPSPLIB with the average project delay (APD) as optimisation criterion.

The Traveling Purchaser Problem (TPP) is a generalization of the Traveling Salesman Problem that can be used to model various real-world applications. Several exact and heuristic approaches were proposed to solving the TPP (and its variants) in recent decades, including an Ant Colony Optimization–based one. We propose an alternative implementation based on the MAX–MIN Ant System and show that it is very competitive when solving the Uncapacitated TPP (U-TPP). When considering the U-TPP instances from a well-known repository by Laporte the proposed algorithm was able to find the optimum to all of the 89 closed instances in an average time of fewer than 3 s. Similarly, it was able to reach the best-known solutions for the 49 out of 51 open instances, while finding a new best solution for one instance.

The problem of knowledge bases verification is now recognized as an important problem in the knowledge engineering. In this article the selected verification issues were considered and the new, experimental version of the verification module of KBExplorer system was introduced. The verification module was implemented as front-end, single page application. The module works on preloaded data, retrieved from back-end server via REST API. The research described in this work are focused on the experimental evaluation of effectiveness of the verification algorithms implemented in JavaScript. The work presents the outline of proposed verification algorithms. Experiments were conducted on real-world knowledge bases with a relatively large number of rules. The results obtained for three rule bases, three hardware configuration and the web-browsed are compared and some conclusions are drawn.

The article presents a description of proposed methods for knowledge representation. Both descriptive and visualizive approaches are included. It uses CluVis software with rules clustering and visualization implementation. The agglomerative hierarchical clustering algorithm is used to generate the rule clusters. The resulting groups are visualized using the tree maps method. Generated rule clusters are labelled with representatives using various methods: threshold, lower and upper approximation as well as weighted method. The experiments have been performed on real rule-based knowledge base from medical domain. The paper contains the analysis of the influence of different representative methods on the representation of knowledge bases and the efficiency of inference processes.

In this paper the topic of clustering and searching through clusters generated from real-world knowledge bases is discussed. Authors analyze three methods of cluster’s representatives creation, focusing on their advantages and flaws. What is more the authors introduce a concept of forward-chaining inference which uses a density-based DBSCAN algorithm to generate cluster of rules and speed up the whole inference process. The experiments were conducted on real-world knowledge bases with a relatively large number of rules to evaluate the efficiency of the proposed approach.

The article considers the approach in which global decisions are made based on dispersed knowledge. Two approaches that were proposed in the previous papers are considered in this article. One of them consists in direct application of Pawlak conflict analysis model. The second is to use a system with two stages of the process of local models coalitions creation. In both systems, the same fusion methods are used, in order to combine local decisions that were generated by coalitions. The following methods are applied: maximum rule, minimum rule, median rule, sum rule and product rule. The obtained results are compared and conclusions are drawn.

The aim of this paper is to introduce an automatic assigning algorithm for a smart building system. Although the addition of a new device to a system always needs some actions such as assigning of an address and setting up specified parameters, none of the current solutions have automated this problem. This paper will describe a Home Automation System (HAuSy) framework with an automatic wireless node addressability process for use in large buildings, such as smart hotels. An optimized algorithm to automatically assign a large number of nodes in a smart building environment will be presented. This algorithm will be benchmarked in simulated tests. Finally, the benefits and possible extensions of this algorithm are discussed at the end of this article.

This article presents a computer expert system that evaluates an air-strike threat. The presented system uses fuzzy methods and it has been created based on the expert knowledge of a given problematic. Our system is able to calculate the threat level of a monitored aircraft for a protected object in real time. It evaluates the inter-relationships between air attack elements and protected objects that are attacked or where it can be assumed that they will be attacked. The results of this research have been checked and verified by experts in military tactics.

In this article, we analyze the effectiveness of the telemarketing campaign developed by the Portuguese bank. The main goal of this analysis was to estimate, is there existing ant colony algorithms are capable of building classifiers that lead to increasing the effectiveness of the telemarketing campaign. An additional question was related to the problem of adjusting the whole campaign to the actual needs of clients. Presented data include 17 attributes, including information about the efficiency of carried out conversations related to the bank deposit offer. The analysis presented in this article was developed on the basis of algorithms used for the decision trees construction such as CART and C4.5. As a result, a prediction allowing to estimate the result of the telemarketing conversation with a client was made. Conducted experiments allowed for the comparison of different classifiers. The comparison was made on the basis of different measures of classification efficiency. It is especially important in the case of the real-world data, where cardinality of decision classes is uneven. Conducted experiments allowed for the comparison of different classifiers. Initial evaluation confirms, that such an approach could be efficiently used for the dynamic data sets, like streams.

In this article, we propose a novel approach to transforming financial time-series values into the symbolic representation based on value changes. Such approach seems to have a few advantages over existing approaches, while one of the most obvious is noise reduction in the data and possibility to find patterns which are universal for investigating different currency pairs. To achieve the goal we introduce the preprocessing method allowing the initial data transformation. We also define a text-based similarity measure which can be used as an alternative for methods allowing to find exact patterns in the historical data.The proposed approach is experimentally verified on 10 different currency pairs, each covering approximately period of 10 years.

This paper was proposed a new method for suggesting creating folders in users mailboxes by using Ant Colony Optimization algorithms and Social Networks Analysis. The aim of this paper is to create a mechanism to predict new folders in automatic classification email messages to folders in the mailboxes. The proposed algorithm uses the elements of Social Networks Analysis used to determine the groups of users who have a similar folder structure in mailboxes, on the basis of which the mechanism suggest new folders for the users. The operation of the proposed method has been tested on a public Enron E-mail Dataset.

Collective intelligence represented as sentiment extracted from social media mining found applications in various areas. Numerous studies involving machine learning modelling have demonstrated that such sentiment information may or may not have predictive power on the stock market trend. This research investigates the predictive information of sentiment regarding the Energy Select Sector related XLE index and of its constituents, on the index and its volatility, based on a novel robust machine learning approach. While we demonstrate that sentiment does not have any impact on any of the trend prediction scenarios investigated here related to XLE and its constituents, the sentiment’s impact on volatility predictions is significant. The proposed volatility prediction modelling approach, based on Jordan and Elman recurrent neural networks, demonstrates that the addition of sentiment or sentiment moment reduces the prediction root mean square error (RMSE) to about one third. The experiments we conducted also demonstrate that the addition of sentiment reduces the RMSE for 24 out of the 36 stocks/constituents, representing 87.9% of the index weight. This is the first study in the literature relating to the prediction of the market trend or the volatility based on an index and its constituents’ sentiment.

Android is a most popular mobile-based operating system with billions of active users, which has encouraged hackers and cyber-criminals to push the malware into this operating system. Accordingly, extensive research has been conducted on malware analysis and detection for Android in recent years; and Android has developed and implemented numerous security controls to deal with the problems, including unique ID (UID) for each application, system permissions, and its distribution platform Google Play. In this paper, we evaluate four tree-based machine learning algorithms for detecting Android malware in conjunction with a substring-based feature selection method for the classifiers. In the experiments 11,120 apps of the DREBIN dataset were used where 5,560 contain malware samples and the rest are benign. It is found that the Random Forest classifier outperforms the best previously reported result (around 94% accuracy, obtained by SVM) with 97.24% accuracy, and thus provides a strong basis for building effective tools for Android malware detection.

When we collect data, usually they consist of small samples with missing values. As a consequence of this flaw, the data analysis becomes less effective. Almost all algorithms for statistical data analysis need a complete data set. In data preprocessing, we have to deal with missing values. Some well-known methods for filling missing values are: Mean, K-nearest neighbours (kNN), fuzzy K-means (FKM), etc. There are quite a lot of R packages offering the imputation of missing values, but sometimes its hard to find the appropriate algorithm for a particular dataset. When we have to deal with large datasets sometimes, these known methods cannot work as supposed because they need too much memory to perform their operations. This paper provides an overview of a considerable dataset imputation by applying three different algorithms. A comparison was performed using three different algorithms under a missing completely at random (MCAR) assumption, and based on the evaluation criteria: Root mean squared error (RMSE). The experiment results show that Random Forest algorithm can be quite useful for missing values imputation.

The paper deals with the intergeneration municipality project on the development of computer literacy in the elderly people with focus on the management of conducted courses. The aim of the paper is to design and propose a model of the project management. The sub-goal is to identify which computer skills attendees of computer literacy courses have and which computer skills they would like to gain or develop. The next issue is to find whether pupils that give lectures to the elderly people have necessary computer skills. The main purpose of the research is to gain current data needed for the updating phase of the long-term project. Key data and findings are collected from own questionnaire investigation and interviews with the municipality, directors of elementary schools, teachers, pupils and attendees which were conducted during computer courses for elderly people. Results and recommendations will be used for the next run of courses that are organized twice a year by municipality in cooperation with primary schools. The biggest ratio of participants can use the Internet followed by email and Skype. MS Excel and Instagram represent the smallest ratio of applications that pupils can use. All other followed applications are used by more than 85%. Results are important for the project team so that they could monitor and evaluate the progress of the project and work out its updated phase.

Financial literacy now appears to be a key component of a successful national financial education strategy. Financial education, financial consumer protection and financial inclusion are the goals of top country policies. For these reasons, it is necessary to assess the levels of financial literacy in each country. The emphasis on financial literacy is already devoted to small children and includes the entire population. The aim of the paper is the OECD international comparison, which focuses on three key areas: financial knowledge, attitudes and behavior. Based on this questionnaire, the situation in the Czech Republic was analyzed. Subsequently, web portals of Czech state institutions (Czech National Bank, The Ministry of Finance of the Czech Republic, Czech Banking Association) focused on financial education were presented. The individual portals are evaluated and checked to see if they meet the standards of financial literacy. These portals are useful for teaching using blended learning which is gaining increasing attention in the Czech Republic. Primary and secondary sources were used in the paper.

Use of new technologies and students skills and competencies in the connection with hybrid learning are solved in the article. Authors focus on the students in the Czech Republic. Students use modern technologies not only for communication and leisure time but also in the process of education that is one of the element of SMART cities. The results of the statistics show that the use of modern technologies has an increasing trend not only in the Czech Republic. The age groups show that with increasing age there is a declining trend for the use of modern technologies. Almost 90% of students can use text processor. Table processor is used by 80.1% of students and presentation software is less used tool in this group by 67.2% of students. Internet is the most often used by students. Results from network readiness index that are presented in the article corresponds with results focused on the computer skills and usage of technologies by students.

Authors have investigated battery electric road vehicles and their possible impacts on the environment and on the Hungarian and Slovakian energy sector. In this article authors investigated the effect of electric market on the emission of electric cars. After analyzing the tendencies of electric market the tendencies of plug.in electric vehicles were analyzed. As a conclusion authors have showed the significant drop of emission caused by electric vehicles in Hungary and Slovakia. The main finding are from this article that battery electric cars could significantly decrease the emission of transport sector only and if only that the appropriate energy sector support this.

Irrepressible increase in human population brings out the necessity of new living spaces across the world. Due to the lack of lands for building construction, floating city concept becomes current among considerable number of countries. Moreover, conforming to rising sea levels and climate change, precautions should be urgently taken by companies, researchers and designers. Along with the technological developments, it is now inevitable to construct special settlements on the sea. In this paper, we are gathering current concepts and approaches to lead the scientific era to make more researches for the floating city projects. Meanwhile self-sufficiency as a concept is not studied in the literature, the use of renewable resources gains importance for sustainable buildings and constructions on the sea. Therefore, since low budget is the main purpose of every construction, material usage should also be taken into consideration for the floating settlements in general. Furthermore, this paper presents the applications, research and development of urban city designs which have been studied so far in the literature, and they could be adapted to the design phase of floating city projects. Therefore, most of the floating city projects have novel approaches and concepts, yet further research is still needed for making actual project implementations.

Buildings around the world have been confronting too much energy consumption. On that account, architectural aspects have an important role to minimize this energy consumption. Overheating has become a current topic especially in summer with hot climates, since the use of large glazed facades have become prevalent. Thus, shading devices should be considered in the early stage of the design process to overcome this task. Since the design of a new shading device is a complicated architectural design problem, the presented problem can be tackled with real-parameter multi-objective optimization. The well-known and fast Non-Dominated Sorting Genetic Algorithm II so called NSGA-II was used to solve this complex design problem and identify alternative solutions to decision makers. An implementation of the method is presented, focusing on louvers design integrated to an office building in a hot-dry climate region.

Purpose. Smart technologies are now being integrated into many places and buildings. Different technologies use their own cloud services, which are suitable for use in a smart home but whose use might be problematic in a building with a higher density of persons, such as a hotel. The purpose of this research is to design and create an architecture for a smart control system that can be used in a hotel environment.Methodology. The current trends of using different cloud services in the smart environment are discussed, and the pros and cons of these solutions are outlined. The functionality of our smart control system is based on possible smart hotel scenarios.Findings. The use of different cloud services in the smart hotel environment might be very demanding for the connection resources. There are also issues with the apps/devices that control the systems because they are designed to control smart homes and they are not designed for hotel rooms. A regular change of user might also be problematic. Based on these findings, our smart control system has been designed with a three layer architecture.Value. The value of this research is to propose an architecture for a smart system that can be used in a smart hotel environment. The designed system focuses on privacy, the possibility of connecting many rooms that can be controlled independently, and the speed of the system’s responses.

An endoscopy is a strategy in which a specialist utilizes specific instruments to see and work on the inward vessels and organs of the body. This paper expects to predict the abnormalities and diseases in the Gastro-Intestinal Tract, utilizing multimedia data acquired from endoscopy. Deep Analysis of GI tract pictures can foresee diseases and abnormalities, in its early stages and accordingly spare human lives. In this paper, a novel ensemble method is presented, where texture and deep learning features are integrated to improve the prediction of the abnormalities in the GI tract e.g. Peptic ulcer disease. Multimedia content analysis (to extricate data from the visual information) and machine learning (for classification) have been explored. Deep learning has additionally been joined by means of Transfer learning. Medieval Benchmarking Initiative for Multimedia Evaluation provided the dataset, which includes 8000 pictures. The data is gathered from conventional colonoscopy process. Using logistic regression and ensemble of different extracted features, 83% accuracy and a F1 score of 0.821 is achieved on testing sample. The proposed approach is compared with several state-of-the-art methods and results have indicated significant performance gains when compared with other approaches.

In a clinical practice of the orthopedics, the articular cartilage assessment is one of the major clinical procedures serving as a predictor of the future cartilage loss development. The early stage of the cartilage osteoarthritis is badly observable from the native MR records due to weak contrast between the physiological cartilage and the osteoarthritic spots. Therefore, the cartilage regional modeling would reliably differentiate the physiological cartilage from the early cartilage deterioration, and can serve as an effective clinical tool. In a comparison with the conventional segmentation methods based on the hard thresholding, the soft fuzzy thresholding based on the histogram separation into segmentation classes via the fuzzy triangular functions represents a sensitive regional segmentation even in the non-contrast environment. We have proposed the soft segmentation where the fuzzy sets are driven by the ABC genetic algorithm to optimal fuzzy class’s distribution regarding the knee tissues characteristics. Consequently, the spatial aggregation is employed to taking advantage the spatial dependences which allows for modification the original fuzzy membership function. This procedure ensures the correct pixel’s classification especially when the noise pixels are present. Such multiregional segmentation makes a mathematical model well separating the physiological cartilage from the early osteoarthritic spots which are highlighted in the model.

GlioBastoma Multiforme (GBM) is an aggressive primary brain tumor characterized by a heterogeneous cell population that is genetically unstable and resistant to chemotherapy. Indeed, despite advances in medicine, patients diagnosed with GBM have a median survival of just one year. Magnetic Resonance Imaging (MRI) is the most widely used imaging technique for determining the location and size of brain tumors. Indisputably, this technique plays a major role in the diagnosis, treatment planning, and prognosis of GBM. Therefore, this study proposes a new Case Based Reasoning approach to problem solving that attempts to predict a patient’s GBM volume after five months of treatment based on features extracted from MR images and patient attributes such as age, gender, and type of treatment.

Beautiful paintings can be approximated with remarkably decent quality using a finite list of translucent polygons, each consisting of a finite number of points, and initialized with random color and coordinates. The polygons evolve by repeatedly mutating their color and coordinates until the resulting mutant satisfies some selection criteria for the next generation. In the end, an approximation of the given image is achieved with a good precision given the restriction that the number of polygons and the number of points per polygon are limited. Since its appearance in 2008 under the name “Evolution of Mona Lisa”, researchers’ interest toward it has decreased despite its initial popularity, which can be partially explained with the lack of a formal publication. In this paper, we describe an efficient natural selection strategy inspired by simulated annealing that, when compared to the existing method, yields better results in every experiment that we conducted. Moreover, this may serve as the first formal introduction to this problem and motivate further research on the topic.

Content-based video indexing is still a very intensively developed area of research in computer science. Most frequently the first stages of content recognition are the temporal segmentation and the detection of video structure. The analyses and the observations of different genre of videos confirm that the edition of videos and the video structures significantly depend on the video genre. On the other hand many processes will be better performed if the genre of video is known and the parameters of processes are adequate to the video genre. The paper presents the tests in the AVI Indexer showing that the genre of a video edited in a standard way and typical for a given video genre can be detected only on the basis of the analysis of sequences of the shot lengths.