Intelligent Systems'2014

This paper deals with intelligent counting, i.e. counting performed under imprecision, fuzziness of information about the objects of counting. Formally, this collapses to counting in fuzzy sets. We will show that the presented methods of intelligent counting are human-consistent, and reflect and formalize real, human counting procedures. Other applications of intelligent counting in intelligent systems and decision support, including questions of similarity measures and time series analysis, will also be outlined or mentioned.

When making recommendations based on the aggregated correlations both their absolute values and signs are meaningful and important. This is the reason why traditional aggregation operators, like OWA functions, may not be satisfactory. Therefore, a generalization of OWA operators which might be useful in aggregating a bipolar information is proposed and examined.

The averaging aggregation operators are defined and some interesting properties are derived. Moreover, we have extended concave and convex property. The main results concerning aggregation of generalized quasiconcave and quasiconvex functions are presented and some properties of aggregation operators are derived and discussed. We study the class of concavity and convexity of two variable aggregation operators that preserve these properties.

Uninorms are an important generalization of triangular norms and triangular conorms. Uninorms allow the neutral element to lie anywhere in the unit interval rather than at zero or one as in the case of a

t

-norm and a

t

-conorm.

Since interval valued fuzzy sets, Atanassov’s intuitionistic fuzzy sets and

L

I

-fuzzy sets are equipollent, therefore in this paper we describe a generalization of uninorms on

L

I

. For example, we describe the structure of uninorms, discuss the possible values of the zero element for uninorms and of the neutral element, especially for decomposable uninorms.

This paper is mainly devoted to solving the functional equations of modularity of special class of aggregation operators with 2-neutral elements. Our investigations are motivated by modular logical connectives and their generalizations used in fuzzy set theory e.g. triangular norms, conorms, uninorms and nullnorms. In this work the modularity of two binary operations from the family of 2-uninorms (

$\textbf{U}_{k(e,f)}$

) which generalizes both uninorms and nullnorms is considered. In particular, all of the possible solutions for one of the three classes of these operations depending on the position of its absorbing and neutral elements are characterized.

Linguistic summaries of databases provide quick insight in the stored data and are important facilities for understanding and grasping the meaning of large data collections. This is especially relevant in the context of big data. However, large data collections often suffer from incomplete data, which are in the case of relational databases modelled by so-called null-values. In this paper we propose a novel soft computing technique for measuring the quality of a linguistic summary in the case of missing information. More specifically we describe and illustrate how bipolar satisfaction degrees can be used to model both the validity of the summary and the hesitation about this validity that might be caused due to missing information. The extra information about the hesitation provides the users with a semantically richer description of the summarization results, which is important in view of a correct interpretation.

In this paper we investigate linguistic summaries of the form “

Q

y

’s are

P

”. We consider a case with non-monotonic quantifier, exemplified by

a few

or

about a half

. We propose a method for evaluating the truth value of such summaries.

Some models using metaheuristics based in an “improvement of solutions” procedure, specifically Genetic Algorithms (GA), have been proposed previously to the linguistic summarization of numerical data (LDS). In the present work is proposed a new model for LDS based in Ant Colony Optimization (ACO), a metaheuristic that use a “construction of solution” procedure. Both models are compared in LDS over

creep

data. Results show how the ACO based model overcomes the measures of goodness of the final summary but fails to improve the results of the GA based model in relation to the diversity of the summary. Features of both models are considered to explain the results.

In this paper, we present some interesting findings from the application of our recently developed InterCriteria Decision Making (ICDM) approach to data extracted from the World Economic Forum’s Global Competitiveness Reports for the years 2008–2009 to 2013–2014 for the current 28 Member States of the European Union. The developed approach which employs the apparatuses of index matrices and intuitionistic fuzzy sets is designed to produce from an existing index matrix with multiobject multicriteria evaluations a new index matrix that contains intuitionistic fuzzy pairs with the correlations revealed to exist in between the set of evaluation criteria, which are not obligatory there ‘by design’ of the WEF’s methodology but exist due to the integral, organic nature of economic data. Here, we analyse the data from the six-year period within a reasonably chosen intervals for the thresholds of the intuitionistic fuzzy functions of membership and non-membership, and make a series of observations about the current trends in the factors of competitiveness of the European Union. The whole research and the conclusions derived are in line with WEF’s address to state policy makers to identify and strengthen the transformative forces that will drive future economic growth.

In this paper, we continue our investigations of the newly developed InterCriteria Decision Making (ICDM) approach with considerations about the more appropriate choice of the employed intuitionistic fuzzy threshold values. In theoretical aspect, our aim is to identify the relations between the thresholds of inclusion of new elements to the set of strictly correlating criteria and the numbers of correlating pairs of criteria thus formed. We illustrate the findings with data extracted from the World Economic Forum’s Global Competitiveness Reports for the years 2008–2009 to 2013–2014 for the current 28 Member States of the European Union. The study of the findings from the considered six-year period involves trend analysis and computation of two approximating functions: a linear function and a polynomial function of 6

th

order. The per-year trend analysis of each of the 12 criteria, called ‘pillars of competitiveness’ in the WEF’s GCR methodology, gives an opportunity to prognosticate their values for the forthcoming year 2014–2015.

The paper deals with cooperation problems in the case of two parties having different sets of criteria measuring their payoffs. Using ideas of the game theory, a mathematical model describing multicriteria bargaining problem is formulated. In the paper an interactive procedure supporting multicriteria analysis and aiding consensus seeking is presented which can be implemented in a computer-based system. According to the procedure the system supports multicriteria analysis made by the parties and generates mediation proposals. The mediation proposals are derived on the basis of the original solution to the multicriteria problem. The solution expresses preferences of the decision makers. It generalizes the classic Raifa solution concept on the multicriteria case.

This article describes the use of the bees algorithm for optimization of the multi-object structure of a welded beam. This problem has been used as a benchmark of the bees algorithm. This article presents an approach to improve the ABC algorithm criteria used for the WBDP multicriteria optimization. The contents of the article show the analysis of the problem, conducted based on the mathematical calculations of the beam dimensions. The further part constitutes a comparison of the received results with the results of the ABC operation. Based on the standards, an additional criterion was established, which verifies the correctness of the results.

L. Ciungu and B. Riečan proved in [2] that any real state on IF-sets can be represented by integrals in sense that

$$m{\left(\left(\mu_A, \nu_A\right)\right)}=\int{\mu_A\mathrm{d}P}+\alpha\left(1-\int{\left(\mu_A+\nu_A\right)\mathrm{d}Q}\right).$$

However the formulation is unappropriate for general case with values from arbitrary Riesz space. This article shows that only small change in formulation make it appropriate for the general case.

Because Riesz spaces have similar structure it is natural question: Does this equality hold in general Riesz space? The answer is probably no because in general case it can be possible that there exist elements

u

,

v

of Riesz space such that 0 ≤ 

u

 ≤ 

v

, but there is no

α

 ∈ ℝ such that

u

 = 

αv

. However this inconvenience wanishes if we rewrite this by the folowing way: For any real state on IF-set there exist measures

P

,

Q

such that

$$m{\left(\left(\mu_A, \nu_A\right)\right)}=\int{\mu_A\mathrm{d}P}+\int{\left(1-\mu_A-\nu_A\right)\mathrm{d}Q}.$$

This formulation is appropriate for general case.

It is well known that the set

F

of

IF

-sets can be embedded to an

MV

-algebra

M

. In the contribution to any finitely additive state

m

on

F

there is constructed a finitely additive state

$\bar m$

on

M

which is an extension of

m

.

In [13] any finitely additive

IF

-state has been embedded to some

MV

-algebra. Using this result we embedde any

IF

-state to an

MV

-

σ

-algebra.

In previous research the differential calculus for the functions defined on

IF

-sets was studied. Then there was the natural question if it is possible to define also definitive integral on this structure. In this paper the properties of definitive integral defined on

IF

-sets are studied.

The central issue of inference validity, that guarantees the correctness of reasoning and thus that of derived knowledge, depends on the definition of both implication operator and tautology. This paper studies the

modus ponens

validity in the case of Intuitionistic Fuzzy logic, in an experimental framework: considering 18 classical implication operators, it shows that validity usually does not hold for the classical definition of Intuitionistic Fuzzy tautology. It proposes two alternative, more constrained, tautology definitions, studying them with the same protocol, showing they make it possible to decrease the number of invalid implication operators.

In this paper, we introduce specific types of intuitionistic fuzzy sets, inspired by the multi-dimensional intuitionistic fuzzy sets and the General Type-2 fuzzy sets. The newly proposed sets extend the opportunities of the General Type-2 fuzzy sets when modelling of particular types of uncertainty. Short comparison between concepts of interval type-2 fuzzy sets and intuitionistic fuzzy sets is given.

Time is a basic aspect in any field, as factors evolve over time and influence the progression of any procedure. This work proposes an integrated approach to developing Timed Fuzzy Cognitive Maps (T-FCMs), an extension of FCMs that can handle uncertainty to infer a result. TFCMs take into consideration the time evolution of any procedure and permit the production of intermediate results and the influence of exterior parameters. It described the proposed method to develop T-FCM and then T-FCM is applied to develop a Medical Decision Support System.

In the paper we study an issue of building tools for accessing and processing knowledge. Such tools are fundamental facets of user friendly human-machine communication. An efficient understanding of data becomes of paramount relevance when dealing with a wealth of human-system interaction and communication. The underlying objective of this paper is to elaborate concepts of aforementioned tools based on automatic data understanding. The syntactic and semantic facets of data integrated in frames of granularity act as fundamental mechanism for data structuring in a form of granular cognitive maps. Linguistic approach to data structuring allows for establishing a suitable perspective at the problem at hand where knowledge structures need to be accessed and processed. Rather than embarking on the formal framework, our intent is to illustrate a realization of the paradigm in the realm of music information, its processing and understanding.

This study is concerned with an issue of automatic data understanding being treated as a fundamental tool used in cognitive maps creation. The underlying objective of this paper is to elaborate on a paradigm of automatic data understanding. We highlight the syntactic and semantic aspects of data. Granular Computing and information granules play a central role in all processes of data understanding by facilitating establishing a suitable perspective at the problem at hand where the data need to be looked at. Our intent is to illustrate a realization of the paradigm in the realm of music information, its processing and understanding, rather than embarking on the formal framework.

We present a general framework for the synthesis of the constraints under which the selected properties hold in a class of models with discrete transitions, together with Boolean encoding - based method of implementing the theory. We introduce notions of parametric image and preimage, and show how to use them to build fixed-point algorithms for parametric model checking of reachability and deadlock freedom. An outline of how the ideas shown in this paper were specialized for an extension of Computation Tree Logic is given together with some experimental results.

In this work, we identify the main problems in which methodology of creating multi-game playing programs differs from single-game playing programs. The multi-game framework chosen in this comparison is General Game Playing, which was proposed at Stanford University in 2005, since it defines current state-of-the-art trends in the area. Based on the results from the International General Game Playing Competitions and additional results of our agent named MINI-Player we conclude on what defines a successful player. The most successful players have been using a minimal knowledge and a mechanism called Monte Carlo Tree-Search, which is simulation-based and self-improving over time.

In this paper, we introduce a flexible consensus reaching process when agents evaluate the alternatives through linguistic expressions formed by a linguistic term, when they are confident on their opinions, or by several consecutive linguistic terms, when they hesitate. Taking into account an appropriate metric on the set of linguistic expressions and an aggregation function, a degree of consensus is obtained for each alternative. An overall degree of consensus is obtained by combining the degrees of consensus on the alternatives by means of an aggregation function. If that overall degree of consensus reaches a previously fixed threshold, then a voting system is applied. Otherwise, a moderator initiates a consensus reaching process by inviting some agents to modify their assessments in order to increase the consensus.

In this paper we introduce a decision system for supporting the ranking of patents carried on by a group of experts in a multiple criteria fuzzy environment. The process starts with the creation of three value scenarios for each considered patent by each expert which are then used for the construction of individual fuzzy pay-off distribution functions for the patent value, here represented with triangular fuzzy numbers. Then, for each expert a TOPSIS matrix is estimated assuming that the scores are linguistically expressed due to the vagueness of individual judgments. We assume that the criteria are represented by the first three possibilistic moments of the pay-off distribution function and by a set of “strategic” attributes that describe different relevant aspects of the patents under analysis. A novel consensus modeling mechanism is then introduced to determine a coalition of experts whose TOPSIS-based evaluations are close enough. Finally, the coalition-based group TOPSIS ranking of patents is determined.

The main goal of this paper is to support reaching a consensus type solution in a group decision making problem. In this context, we present a new model which assists the support system which is strongly integrated with our consensus reaching model. It is based on a role of a moderator who helps agents (individuals), by rational argument, persuasion, etc. change their opinions and towards a higher agreement within the entire group. As in our previous works, the consensus degree determines the agreement among most of (important) agents as to most of (relevant) options. Information about the current state of agreement and the main obstacles in reaching consensus are provided in a human consistent, hence easy to use form, by linguistic data summaries that can be derived, for instance, via natural language generation (NLG). In this paper, we extend the consensus evaluation and reaching model by carrying out various consensus reaching scenarios depending on a context of the process and considering different natures of a group of agents involved. Though not each discussion requires the involvement of each member, which may be time consuming, it may often be necessary to avoid the accounting for group member’s interests or emotional needs of agents. Therefore, to cope with these types of scenarios, we propose here an efficiently focused use of additional linguistic consensus indicators to provide the moderator with appropriate mechanisms for guiding the decision makers towards a higher degree of consensus. Finally, we present an illustrative implementation and numerical evaluation of various attitudes of the moderator’s actions in the model proposed.

This paper presents the design and implementation of a collaborative multi-agent system for automated negotiation processes occurring in dynamic networked environments. The focus is given to a negotiation mechanism required at business level. The mechanism is achieved by software agents within a collaborative working environment in the special case of notified transient situations. Automated negotiation is one of the most common approaches used to make decisions and manage disputes between computational entities leading them to optimal agreements.

The proposed solution is based on a multi-agent system architecture that applies rule-based negotiation at business level. This architecture has been tested in the case of on-line auctions in distributed networks.

In this paper, an extension of the concept of an index matrix is introduced and some of its basic properties are studied. Operations, relations and operators are defined over the new object.

In this article, some possible interpretations of the computational model based on discrete fuzzy numbers are given. In particular, some advantages of this model based on the aggregation process as well as on a greater flexibilization of the linguistic expressions are analysed. Finally, a fuzzy decision making model based on this kind on fuzzy subsets is proposed.

The uncertainty theory solves problems with uncertain data. Often to perform arithmetic operations on uncertain data, the calculations on intervals are necessary. Interval arithmetic uses traditional mathematics in the calculations on intervals. There are many methods that solve the problems of uncertain data presented in the form of intervals, each of them can give in some cases different results. The most known arithmetic, often used by scientists in calculations is Moore interval arithmetic. The article presents a comparison of Moore interval arithmetic and multidimensional RDM interval arithmetic. Also, in both Moore and RDM arithmetic the basic operations and their properties are described. Solved examples show that the results obtained using the RDM arithmetic are multidimensional while Moore arithmetic gives one-dimensional solution.

In the paper properties of interval-valued fuzzy preference relations are considered and preservation of a preference property by some operations, including lattice operations, the converse and the complement relations are studied. The concept of a preference relation presented here is a generalization of the concept of crisp preference relations. Moreover, weak properties of interval-valued fuzzy relations, namely reflexivity, irreflexivity, connectedness, asymmetry, antisymmetry, transitivity, and moderate transitivity are defined. Furthermore, the assumptions under which interval-valued fuzzy preference relations fulfil the mentioned properties are proposed.

Database systems contain data representing properties of real-life objects or concepts. Many of these data represent time indications and such time indications are often subject to imperfections. Although several existing proposals deal with the modeling of uncertainty or vagueness in time indications in database systems, only a few of them summarily examine the interpretation and semantics of such imperfections. The work presented in this paper starts at a more thorough examination of the semantics and modeling of uncertainty or vagueness in time intervals in database systems and presents methods to model combinations of uncertainty and vagueness in time intervals in database systems, based on examinations of their requisite interpretations.

The framework of Approximate Tolerance Geometry (ATG) has been proposed in [1] as an approach to handling large and heterogeneous imperfections in geometric data in vector-based geographic information systems. Here, different types of positional error can often only be subsumed as possibilistic location constraints. The application of the ATG framework to a classical geometry provides a calculus for the propagation of this error type in geometric reasoning. As a first step towards an implementation of an ATG geometry, the paper applies the framework to the geometric equality relation. It thereby lays the basis for the application of ATG to the other axioms of classical geometry.

The paper presents the estimators of three relations: equivalence, tolerance and preference in a finite set on the basis of multiple pairwise comparisons, disturbed by random errors; they have been developed by the author. The estimators can rest on: binary (qualitative), multivalent (quantitative) and combined comparisons. The estimates are obtained on the basis of discrete programming problems. They require weak assumptions about distributions of comparisons errors, especially allow non-zero expected values. The estimators have good statistical properties, in particular are consistent. The estimates can be verified using statistical tests. The paper summarizes briefly the results obtained lastly by the author.

Intelligent virtual agents function in dynamic, uncertain, and uncontrolled environments, and animating them is a chaotic and error-prone task which demands high-level behavioral controllers to be able to adapt to failures at lower levels of the system. On the other hand, the conditions in which space robotic systems such as spacecraft and rovers operate, inspire by necessity, the development of robust and adaptive control software. In this paper, we propose a generic architecture for developing autonomous virtual agents that let them to illustrate robust deliberative and reactive behaviors, concurrently. This architecture is inspired by onboard autonomous frameworks utilized in interplanetary missions. The proposed architecture is implemented within a discrete-event simulated world to evaluate its deliberative and reactive behaviors. Evaluation results suggest that the architecture supports both behaviors, consistently.

Paper proposes a novel solution to a job scheduling problem in the Cloud Computing systems. The goal of this scheme is allocating a limited quantity of resources to a specific number of jobs minimizing their execution failure probability and completion time. It employs the Pareto dominance concept implemented at the client level. To select the best scheduling strategies from the Pareto frontier and construct a global scheduling solution we develop decision-making mechanisms based on the game-theoretic model of Spatial Prisoner’s Dilemma and realized by selfish agents operating in the two-dimensional Cellular Automata space. Their behavior is conditioned by objectives of the various entities involved in the scheduling process and driven towards a Nash equilibrium solution by the employed social welfare criteria. The related results show the effectiveness and scalability of this scheme in the presence of a large number of jobs and resources involved in the scheduling process.

The problem of practical model checking Alternating-time Temporal Logic (ATL) formulae under imperfect information and imperfect recall is considered. This is done by synthesis and subsequent verification of

strategies

, until a good one is found. To reduce the complexity of the problem we define an equivalence relation on strategies. Then an algorithm for model checking a class of modal properties with a single coalitional modality is presented, which utilises the observation that there is no need to verify more than one strategy from an equivalence class. The experimental results of the approach are also discussed.

The paper provides a general description of a model for context-aware agent architecture (C3A) and first steps in AjTempura creation vie C3A model. The approach adopts the definition of context and context-awareness given by Dey. The C3A model aims at creating of smart virtual spaces. The applicability of the model is demonstrated by development of an agent-oriented application.

In this paper we propose a Neutral Mutation Operator (NMO) for Grammatical Evolution (GE). This novel operator is inspired by GE’s ability to create genetic diversity without causing changes in the phenotype. Neutral mutation happens naturally in the algorithm; however, forcing such changes increases success rates in symbolic regression problems profoundly with very low additional CPU and memory cost. By exploiting the genotype-phenotype mapping, this additional mutation operator allows the algorithm to explore the search space more efficiently by keeping constant genetic diversity in the population which increases the mutation potential. The NMO can be applied in combination with any other genetic operator or even different search algorithms (e.g. Differential Evolution or Particle Swarm Optimization) for GE and works especially well in small populations and larger problems.

Modern optimization has in its disposal an immense variety of heuristic algorithms which can effectively deal with both continuous and combinatorial optimization problems. Recent years brought in this area fast development of unconventional methods inspired by phenomena found in nature. Flower Pollination Algorithm based on pollination mechanisms of flowering plants constitutes an example of such technique. The paper presents first a detailed description of this algorithm. Then results of experimental study of its properties for selected benchmark continuous optimization problems are given. Finally, the performance the algorithm is discussed, predominantly in comparison with the well-known Particle Swarm Optimization Algorithm.

Genetic algorithms are widely used for solving the optimization problems. However combinatorial problems are usually hard to solve using genetic algorithms as the chromosomes are very long, what causes the increase of the computational complexity of such solutions. In the previous research, authors proposed the method of efficient search space reduction, which was applied to lower the complexity of random searches. In the current work, a modification of the multi-agent genetic algorithm and its application for solving the single source shortest path problem are proposed. Presented approach is compared to the former implementation of the genetic algorithm. Investigations showed that the genetic diversity of the population in multi-agent genetic algorithm can be successfully measured, which allows to identify the premature convergence.

The Krill Herd Algorithm is the latest heuristic technique to be applied in deriving best solution within various optimization tasks. While there has been a few scientific papers written about this algorithm, none of these have described how its numerous basic parameters impact upon the quality of selected solutions. This paper is intended to contribute towards improving the aforementioned situation, by examining empirically the influence of two parameters of the Krill Herd Algorithm, notably, maximum induced speed and inertia weight. These parameters are related to the effect of the herd movement as induced by individual members. In this paper, the results of a study – based on certain examples obtained from the CEC13 competition – are being presented. They appear to show a relation between these selected two parameters and the convergence of the algorithm for particular benchmark problems. Finally, some concluding remarks, based on the performed numerical studies, are provided.

In this paper, we analyze the impact of hybridization on the Cuckoo Search algorithm as applied in the context of business process mining. Thus, we propose six hybrid variants for the algorithm, as obtained by combining the Cuckoo Search algorithm with genetic, Simulated Annealing, and Tabu Search-based components. These components are integrated into the Cuckoo Search algorithm at the steps that correspond to generating the new business process models. The hybrid algorithm variants proposed have been comparatively evaluated on a set of event logs of different complexities. Our experimental results obtained have been compared with the ones as provided by the state of the art Genetic Miner algorithm.

In this paper, real-time implementation of recently developed direct particle swarm controller for repetitive process is presented. The proposed controller solves the dynamic optimization problem of shaping the control signal in the voltage source inverter. The challenges in real time implementation come from limited sampling period to evaluate a candidate solution. In this paper, the solution of PDPSRC time-distributed calculation is presented. This method can be implemented using digital signal controllers (DSC).

Wireless Sensor Network consists of an enormous number of small disposable sensors which have limited energy. The sensor nodes equipped with limited power sources. Therefore, efficiently utilizing sensor nodes energy can maintain a prolonged network lifetime. This paper proposes an optimized hierarchical routing technique which aims to reduce the energy consumption and prolong network lifetime. In this technique, the selection of optimal cluster heads (CHs) locations is based on Artificial Fish Swarm Algorithm (AFSA). Various behaviors in AFSA such as preying, swarming, and following are applied to select the best locations of CHs. A fitness function is used to compare between these behaviors to select the best CHs. The model developed is simulated in MATLAB. Simulation results show the stability and efficiency of the proposed technique. The results are obtained in terms of number of alive nodes and the energy residual mean value after some communication rounds. To prove the AFSA efficiency of energy consumption, we have compared it to LEACH and PSO. Simulation results show that the proposed method outperforms both LEACH and PSO in terms of first node die (FND) round, total data received by base station, network lifetime, and energy consume per round.

Time is a crucial dimension in many application domains. This paper proposes an intelligent approach to querying temporal databases using fuzzy temporal criteria. Relying on fuzzy temporal Allen relations, a particular class of criteria are studied. First, a query language that supports flexible temporal query is discussed. Then, the architecture and the interface of the system developed are explicitly described. To evaluate intelligently temporal queries, our system is endowed with some reasoning capabilities.

Detection of outliers can identify defects, remove impurities in the data and what is the most important it supports the decision-making processes. In the paper an outlier detection method based on simultaneous indication of outliers by a group of algorithms is proposed. Three well known algorithms: DBSCAN, CLARANS and COF are considered. They are used simultaneously with iteratively chosen input parameters, which finally guarantee stabilization of the number of detected outliers. The research is based on data retrieved from the Internet service allegro.pl, where comments in online auctions are considered as outliers.

In the setting of relational databases, the schema of the database provides a context in which the data should be interpreted. As a consequence, the quality of a relational database depends strongly on the assumption that data fits this context description. In this paper, we investigate the case where the information provided by an attribute value exceeds the framework provided by the schema. It is shown that such an information overflow can have two orthogonal causes: (i) data about multiple attributes are jointly stored as one attribute and (ii) data about multiple tuples are jointly stored as one tuple. Needless to say, such erroneous information storage deteriorates the quality of the database. In this paper, it is investigated how data quality can be improved by a split operator. The major difficulty hereby is to take into account the constraints that are present in a relational database. A generic algorithm is provided and tested on the well-know Cora dataset.

Cloud Storage Service (CSS) is a vital service of cloud computing which relieves the burden of storage management, cost and maintenance. However, Cloud storage introduces new security and privacy challenges that make data owners worry about their data. It is essential to have an auditing service to verify the integrity of outsourced data and to prove to data owners that their data is correctly stored in the Cloud. Recently, many researchers have focused on validating the integrity of outsourced data and proposed various schemes to audit the data stored on CSS. However, most of those schemes deal with static and single copy data files and do not consider data dynamic operations on replicated data. Furthermore, they do not have the facility to repair corrupt data. In this paper, we address these challenging issues and propose a public auditing scheme for multiple-copy outsourced data in CSS. Our scheme achieves better reliability, availability, and scalability by supporting replication and data recovery.

The XPath query language offers a standard for information extraction from HTML documents. Therefore, the DOM tree representation is typically used, which models the hierarchical structure of the document. One of the key aspects of HTML is the separation of data and the structure that is used to represent it. A consequence thereof is that data extraction algorithms usually fail to identify data if the structure of a document is changed. In this paper, it is investigated how a set of tabular oriented XPath queries can be adapted in such a way it deals with modifications in the DOM tree of an HTML document. The basic idea is hereby that if data has already been extracted in the past, it could be used to reconstruct XPath queries that retrieve the data from a different DOM tree. Experimental results show the accuracy of our method.

Useful information is often scattered over multiple sources. Therefore, automatic data integration that guarantees high data quality is extremely important. One of the crucial operations in data integration from different sources is the detection of different representations of the same piece of information (called

coreferent

data) and translation to a common, unified representation. That translation is also known as

value mapping

. However, values mappings are often not explicit i.e. the specific value may be mapped to more than one value. In this paper, we investigate automatic selection method which reduces the set of one-to-many mappings to the set of one-to-one mappings for attributes whose domains are partially ordered and where the given order relation reflects a notion of generality.

Human activity recognition has become a very important research topic, due to its multiple applications in areas such as pervasive computing, surveillance, context-aware computing, ambient assistive living or social robotics. For activity recognition approaches to be properly developed and tested, annotated datasets are a key resource. However, few research works deal with activity annotation methods. In this paper, we describe a knowledge-driven approach to annotate activity datasets automatically. Minimal activity models have to be provided to the tool, which uses a novel algorithm to annotate datasets. Minimal activity models specify action patterns. Those actions are directly linked to sensor activations, which can appear in the dataset in varied orders and with interleaved actions that are not in the pattern itself. The presented algorithm finds those patterns and annotates activities accordingly. Obtained results confirm the reliability and robustness of the approach in several experiments involving noisy and changing activity executions.

The paper shows an example of application of modeling process of diagnostic knowledge with the use of multimodal statement networks. The goal is to present generic approach to modeling complex diagnostic domains by many independent experts. As an object cogeneration power plant with Organic Rankine Cycle is presented. Implementation of diagnostic model was performed using

REx

system that enables among other things knowledge management and the application and preliminary evaluation of diagnostic knowledge of gathered knowledge for the purpose of its further incorporation to diagnostic systems.

The investigated intuitionistic statement networks facilitate designing and development of complex expert systems. Within these networks, statements consist of contents and values. Statement values are based on a concept introduced in intuitionistic fuzzy sets, i.e. they contain independent belief about validity and nonvalidity of information presented by statement contents. The relationships between statements are modeled in the form of a set of necessary as well as sufficient conditions, and these conditions are considered as corresponding inequalities between statement values. The paper introduces gradual forgetting operator of statement values. The networks with such an operator may be used as models of dynamic objects. They allow for non-monotonic reasoning required for monitoring and diagnostic systems.

In this paper the fast learning algorithm of unsupervised adaptive resonance theory ART2 neural network is described. At the beginning of the process the algorithm is illustrated step by step by mathematical formulas and it is shown how individual vector changes its values during the training. The network supports clustering by using competitive learning, normalization and suppression of the noise. At the end of the process we have stable recognition clusters with values according to the vectors.

The learning process algorithm is presented by a Generalized net model.

Two methods for evaluation of the behavior of tokens in Generalized Nets (GNs) are discussed. In the ordinary GNs the evaluations are based on determining whether the characteristics of the tokens meet a predefined criterion. It is shown that in Generalized Nets with Characteristics of the Places (GNCP) the evaluations of the tokens can also be obtained on the basis of the characteristics of the places. The evaluations are obtained in the form of Intuitionistic Fuzzy Pairs (IFPs). Modification of a given GN model is proposed which allows for the evaluation of tokens on the basis of the characteristics of the places to be obtained during the functioning of the net. The modified GN preserves the functioning and the results of the work of the given net.

The paper considers the generalized net description of essential attributes generator which is one of the main part of SEAA method developed for dimensionality reduction of time series. SEAA method (

Symbolic Essential Attributes Approximation

) (Krawczak and Szkatuła, 2014) was developed to reduce the dimensionality of multidimensional time series by generating a new nominal representation of the original data series. The approach is based on the concept of data series envelopes and essential attributes obtained by a multilayer neural network. The considered neural network architecture is based on Cybenko’s theorem and consists of two three-layer neural networks. In this paper the generalized net description of this part of SEAA method is developed in order to show the beauty of generalized nets.

The present paper proposes a developed generalized net model that compares different mathematical models of the process of

E. coli

fed-batch cultivation. A system of four ordinary differential equations describes the main variables of the considered cultivation process, namely, biomass, substrate and acetate. For the purposes of model simulation, we use the software package GN Lite. The GN model compares the simulated performance of the proposed set of mathematical models and selects the best performing one one the basis of a predefined criterion. During the simulation, GN Lite calls the Matlab software environment to solve the process model, presented as a system of nonlinear differential equations, and plots the dynamics of the main process variables’ of the model that has been computed to perform best.

The paper focuses on a radial basis function network initialization. An application of the agent-based population learning algorithm to set RBF networks main parameters including number and locations of centroids is discussed. The main contribution of the paper is proposing and evaluating an agent-based approach to determine unique subset of features independently for each hidden unit. Two versions of the proposed algorithm for selecting values of the RBF networks parameters are considered. The approach is validated experimentally. Advantages and main features of the PLA-based RBF designs are discussed basing on results of the computational experiment.

Combining first-principles knowledge in the form of a mechanistic model with artificial neural networks (ANN) into so-called hybrid models is an attractive approach to improve models for biological wastewater treatment. Although neural networks have been proven to be an effective method in learning nonlinear input-output mappings, their generalization ability is of concern. Generalization ability of ANNs has been improved by combining several ANNs into ensembles, where each ANN provides a solution to the same problem, and the solutions are combined into a single ensemble output. In this paper, a parallel hybrid modeling approach was developed where ANN ensembles were used to improve Activated Sludge Model (ASM) predictions in modeling a pulp mill wastewater treatment plant. This approach was successful in improving the generalization ability of the ASM in modeling three different pollutants. The best performing methods were the cross-validation ensemble and bagging, and in calculating the ensemble output, simple averaging outperformed stacking. It was also shown that the generalization performance was improved by adding more members in the ensemble.

The model in Model Predictive Control (MPC) takes the central place. Therefore, it is very important to find a predictive model that effectively describes the behavior of the system and can easily be incorporated into MPC algorithm. In this paper it is presented implicit Generalized Predictive Controller (GPC) based on Semi Fuzzy Neural Network (SFNN) model. This kind of model works with reduced number of the fuzzy rules and respectively has low computational burden, which make it suitable for real-time applications like predictive controllers. Firstly, to demonstrate the potentials of the SFNN model test experiments with two benchmark chaotic systems - Mackey-Glass and Rossler chaotic time series are studied. After that, the SFNN model is incorporated in GPC and its efficiency is tested by simulation experiments in MATLAB environment to control a Continuous Stirred Tank Reactor (CSTR).

In this paper a comparison of the triangular, Gaussian, trapezoidal and generalized bell membership functions used in the type-2 fuzzy inference systems, which are applied to obtain the type-2 fuzzy weights in the connection between the layers of a neural network. We used two type-2 fuzzy systems that work in the backpropagation learning method with the type-2 fuzzy weight adjustment. We change the type of membership functions of the two type-2 fuzzy systems. The mathematical analysis of the proposed learning method architecture and the adaptation of the type-2 fuzzy weights are presented. The proposed method is based on recent methods that handle weight adaptation and especially fuzzy weights. In this work neural networks with type-2 fuzzy weights are presented. The proposed approach is applied to the case of Mackey-Glass time series prediction.

A new type of a fuzzy model is proposed in this paper. It uses a reduced number of fuzzy rules with respective radial basis activation functions. The optimal number of the rules is defined experimentally and their locations are obtained by clustering or by PSO optimization procedure. All other parameters are also optimized in order to produce the best model. The obtained model is able to work with sparse data in the multidimensional experimental space. As a proof a synthetic example, as well as a real example of a 5-dimensional sparse data have been used. The results obtained show that the PSO optimization of the fuzzy rule locations is a better approach than the clustering algorithm, which utilizes the distribution of the available experimental data.

The paper deals with the problem of ion activity determination for a mixture by means of ion-selective electrodes. Mathematical model of the analysed phenomenon is described by the Nicolsky-Eisenman equation, which relates activities of ions and ion-selective electrode potentials. The equation is strongly nonlinear and, especially in the case of multi-compound assays, the calculation of ion activities becomes a complex task. Application of multilayer perceptron artificial neural networks, which are known as universal approximators, can help to solve this problem. A new proposition of such network has been presented in the paper. The main difference in comparison with the previously proposed networks consists in the input set, which includes not only electrode potentials but also electrode parameters. The good network performance obtained during training has been confirmed by additional tests using measurement results and finally compared with the original as well as the simplified analytical model.

In this paper we propose a new method for classifying uncertain data, modeled as interval-valued fuzzy sets. We develop the notion of an interval-valued prototype-based fuzzy classifier, with the idea of preserving full information including the uncertainty factor about data during the classification process. To this end, the classifier was based on the uncertainty-aware similarity measure, a new concept which we introduce and give an axiomatic definition for. Moreover, an algorithm for determining such a similarity value is proposed, and an application to supporting medical diagnosis is described.

Nearest prototype classifier based on differential evolution algorithm, pool of distances and generalized ordered weighted averaging is introduced. Classifier is based on forming optimal ideal solutions for each class. Besides this also distance measures are optimized for each feature in the data sets to improve recognition process of which class the sample belongs. This leads to a distance vectors, which are now aggregated to a single distance by using generalized weighted averaging (GOWA). In earlier work simple sum was applied in the aggregation process. The classifier is empirically tested with seven data sets. The proposed classifier provided at least comparable accuracy or outperformed the compared classifiers, including the earlier versions of DE classifier and DE classifier with pool of distances.

This paper introduces a new classification method that uses the differential evolution algorithm to feature-wise select, from a pool of distance measures, an optimal distance measure to be used for classification of elements. The distances yielded for each feature by the optimized distance measures are aggregated into an overall distance vector for each element by using OWA based multi-distance aggregation.

We present here a new classifier called an intuitionistic fuzzy decision tree. Performance of the new classifier is verified by analyzing well known benchmark data. The results are compared to some other well known classification algorithms.

The usual characterization of a target set based on an equivalence relation via an information system only involves one instance of data of the information system and one universe. In this paper, I will give an approach to characterize a large target set that would involve several instances and several universes via probabilistic classifiers. First of all, I reduce the large target set to smaller pieces that could be characterized by the universes in hand, and then combine all the characterizations to form a rough set for the large target set. Furthermore, I also devise some ways to choose the optimal reduction(s) and characterization(s) for such large target set.

A target set in a usual characterization is always assumed to be a crisp set and so is the the classifier. In this paper, I give an approach to characterize a target set which is a fuzzy set (or a fuzzy target set) based on fuzzy classifiers induced by data in an information system. I also give a way to compare the efficiency of different information systems in characterizing fuzzy target sets.

This paper proposes a new approach for identifying situations from sensor data by using a perception-based mechanism that has been borrowed from humans: sensation, perception and cognition. The proposed approach is based on two phases: low-level perception and high-level perception. The first one is realized by means of semantic technologies and allows to generate more abstract information from raw sensor data by also considering knowledge about the environment. The second one is realized by means of Fuzzy Formal Concept Analysis and allows to organize and classify abstract information, coming from the first phase, by generating a knowledge representation structure, namely lattice, that can be traversed to obtain information about occurring situation and augment human perception. The work proposes also a sample scenario executed in the context of an early experimentation.

In this paper we present a method to improve the automatic detection of events in short sentences when in the presence of a large number of event classes. Contrary to standard classification techniques such as Support Vector Machines or Random Forest, the proposed Fuzzy Fingerprints method is able to detect all the event classes present in the ACE 2005 Multilingual Corpus, and largely improves the obtained G-Mean value.

Several works have been proposed for both collective and individual trajectory behavior discovery, as flocks, outliers, avoidance, chasing, etc. In this paper we are especially interested in abnormal behaviors of individual trajectories of drivers, and present an algorithm for finding anomalous movements and categorizing levels of driving behavior. Experiments with real trajectory data show that the method correctly finds driving anomalies.

Web software usage used as a vehicle of communication, where sentiments related to different subjects are shared through the web, has had a steep increase in recent years. For example, customers write about characteristics that they like the most, and they dislike, influencing others to use a specific commented service or product. Thus, the content of these comments could be used to address customers’ sentiments in such way that would increase companies’ services quality. However, since the output of such kind of web software is essentially unstructured data it is difficult for company managers to have access to that knowledge, and consequently reason through it accordingly. In this paper, the authors present a framework for customers’ sentiment analysis, which has the ability to automatically acquire knowledge from web software. Thus, it addresses the target of customer’s sentiments with the purpose of supporting companies’ managers in their decision-making to improve their products and services.