One of the main areas of research in Social Network Analysis (SNA) is Link Prediction (LP). The LP problem is useful in understanding the evolution mechanism of social networks, as well as in different applications such as recommendation systems, bioinformatics and marketing. In LP algorithms, prior network information is used to predict future connections in social networks. In this paper, we introduce a multi-wave cellular learning automaton (MWCLA) and use it to solve the LP problem in social networks. This model is a new CLA with a connected structure and a module of LAs in each cell where the neighbours of the cell module are its successors. The MWCLA method uses multiple waves at the same time in the network in order to improve convergence speed as well as accuracy. For predicting links in the social network, multiple waves can be used to consider different aspects of the network. Here we show that the model converges upon a stable and compatible configuration. Compared to some state-of-the-art approaches, MWCLA produces significantly better results when applied to the LP problem.

Learning Automata (LAs) are adaptive decision-making models designed to find an appropriate action in unknown environments. LAs can be classified into two classes: variable structure and fixed structure. To the best of our knowledge, there is no hybrid model based on both of these classes. In this paper, we propose a model that brings together the benefits of both classes of LAs. In the proposed model, called an HLA, the action switching phase of a fixed structure learning automaton is fused with a variable structure learning automaton. Several computer simulations are conducted to study the performance of the proposed model with respect to the total number of rewards and action switching in addition to the convergence rate. The proposed model is compared to both variable structure and fixed structure learning automata, and in most cases, the numerical results demonstrate its superiority. In order to show the applicability of the HLA, a novel adaptive dropout mechanism in deep neural networks was suggested. The results of the simulations show that the proposed mechanism performs better than the simple dropout mechanism with respect to network accuracy.

The fitness evaluation (FE) management has been successfully applied to improve the performance of multi-population methods for dynamic optimisation problems (DOPs). In this work, we extend one of its variants to address DOPs which was recently proposed by the authors. The aim of our proposal is to increase the efficiency of the FE management. To this end, we propose a technique based on hierarchical learning automata that manages FEs at two level: at first level the algorithm decides which population should be executed, and at the second level it specifies the operation that should be performed by the selected population. A detailed experimental analysis shows the effectiveness of our proposal.

Abstract Community structure is an important topological feature of complex networks. Detecting community structure is a highly challenging problem in analyzing complex networks and has great importance in understanding the function and organization of networks. Up until now, numerous algorithms have been proposed for detecting community structure in complex networks. A wide range of these algorithms use the maximization of a quality function called modularity . In this article, three different algorithms, namely, MEM‐net, OMA‐net, and GAOMA‐net, have been proposed for detecting community structure in complex networks. In GAOMA‐net algorithm, which is the main proposed algorithm of this article, the combination of genetic algorithm (GA) and object migrating automata (OMA) has been used. In GAOMA‐net algorithm, the MEM‐net algorithm has been used as a heuristic to generate a portion of the initial population. The experiments on both real‐world and synthetic benchmark networks indicate that GAOMA‐net algorithm is efficient for detecting community structure in complex networks.

The exclusion operator is a key component in separating the search territory of each population in multi-population optimisation algorithms for unconstraint continues dynamic optimisation problems (DOPs) with the aim of maintaining the overall diversity of the population and avoiding redundant search. Although extensively used by the researchers, the role of exclusion has been barely studied in detail. Therefore, in this paper, we solely study the role of exclusion as a part of multi-population methods in DOPs. For this purpose, a comprehensive review of the various exclusion strategies reported in the literature is provided. Four strategies are also introduced to reduce the shortcomings of exclusion operator. Experimental results show that proposed strategies compared to other schemes such as reinitialized midpoint check, hill-valley detection with three checkpoints, and merging information of collided populations have the same or even higher ability to improve the performance of the multi-swarm PSO algorithms in moving peaks benchmark.

A well‐known harmful attack against wireless sensor networks (WSNs) is the Sybil attack. In a Sybil attack, WSN is destabilised by a malicious node which forges a large number of fake identities to disrupt network protocols such as routing, data aggregation, and fair resource allocation. In this study, the authors suggest a new algorithm based on a composition of learning automaton (LA) model and client puzzles theory to identify Sybil nodes in stationary WSNs. In the proposed algorithm, each node sends puzzles to its neighbours periodically during the network lifetime and tries to identify Sybil nodes among them, considering their response time (puzzle solving time). In this algorithm, each node equipped with a LA to reduce the communication and computation overhead of sending and solving puzzles. The proposed algorithm has been simulated using J‐SIM simulator and simulation results have shown that the proposed algorithm can detect 100% of Sybil nodes and the false detection rate is about 5% on average. Also, the performance of the proposed algorithm has been compared to a wellknown neighbour‐based algorithm through experiments and the results have shown that the proposed algorithm is significantly better than this algorithm in terms of detection and false detection rates.

Game theory as one of the most progressive areas in AI in last few years originates from the same root as AI. The unawareness of the other players and their decisions in such incomplete-information problems, make it necessary to use some learning techniques to enhance the decision-making process. Reinforcement learning techniques are studied in this research; regret minimisation (RM) and utility maximisation (UM) techniques as reinforcement learning approaches are widely applied to such scenarios to achieve optimum solutions. In spite of UM, RM techniques enable agents to overcome the shortage of information and enhance the performance of their choices based on regrets, instead of utilities. The idea of merging these two techniques are motivated by iteratively applying UM functions to RM techniques. The main contributions are as follows; first, proposing some novel updating methods based on UM of reinforcement learning approaches for RM; the proposed methods refine RM to accelerate the regret reduction, second, devising different procedures, all relying on RM techniques, in a multi-state predator-prey problem. Third, how the approach, called RMRL, enhances different RM techniques in this problem is studied. Estimated results support the validity of RMRL approach comparing with some UM and RM techniques.

Abstract One of the problems with traditional genetic algorithms (GAs) is premature convergence, which makes them incapable of finding good solutions to the problem. The memetic algorithm (MA) is an extension of the GA. It uses a local search method to either accelerate the discovery of good solutions, for which evolution alone would take too long to discover, or reach solutions that would otherwise be unreachable by evolution or a local search method alone. In this paper, we introduce a new algorithm based on learning automata (LAs) and an MA, and we refer to it as LA‐MA. This algorithm is composed of 2 parts: a genetic section and a memetic section. Evolution is performed in the genetic section, and local search is performed in the memetic section. The basic idea of LA‐MA is to use LAs during the process of searching for solutions in order to create a balance between exploration performed by evolution and exploitation performed by local search. For this purpose, we present a criterion for the estimation of success of the local search at each generation. This criterion is used to calculate the probability of applying the local search to each chromosome. We show that in practice, the proposed probabilistic measure can be estimated reliably. On the basis of the relationship between the genetic section and the memetic section, 3 versions of LA‐MA are introduced. LLA‐MA behaves according to the Lamarckian learning model, BLA‐MA behaves according to the Baldwinian learning model, and HLA‐MA behaves according to both the Baldwinian and Lamarckian learning models. To evaluate the efficiency of these algorithms, they have been used to solve the graph isomorphism problem. The results of computer experimentations have shown that all the proposed algorithms outperform the existing algorithms in terms of quality of solution and rate of convergence.

Firefly algorithm is a swarm based algorithm that can be used for solving optimization problems. This paper proposed an improved fuzzy adaptive firefly algorithm (FAFA). In the proposed FAFA, a fuzzy system is used to adapt Firefly Algorithm’s parameters in order to improve its ability in global and local searches. Also, we used different fireflies initializing intervals and different iteration numbers to show the algorithm capability to find global optima. Results focus on the two case study categories of function optimization (seven benchmark functions) and presented a novel optimal multilevel thresholding approach for histogram-based image segmentation by using proposed FAFA and Otsu method. Evidence indicates that the optimization results of proposed FAFA approach are so better than the standard FA.

Shortest path problem in stochastic graphs has been recently studied in the literature and a number of algorithms has been provided to find it using varieties of learning automata models. However, all these algorithms suffer from two common drawbacks: low speed and lack of a clear termination condition. In this paper, we propose a novel learning automata-based algorithm for this problem which can speed up the process of finding the shortest path using parallelism. For this parallelism, several traverses are initiated, in parallel, from the source node towards the destination node in the graph. During each traverse, required times for traversing from the source node up to any visited node are estimated. The time estimation at each visited node is then given to the learning automaton residing in that node. Using different time estimations provided by different traverses, this learning automaton gradually learns which neighbor of the node is on the shortest path. To set a condition for the termination of the proposed algorithm, we analyze the algorithm using a recently introduced model, Adaptive Stochastic Petri Net (ASPN-LA). The results of this analysis enable us to establish a necessary condition for the termination of the algorithm. To evaluate the performance of the proposed algorithm in comparison to the existing algorithms, we apply it to find the shortest path in six different stochastic graphs. The results of this evaluation indicate that the time required for the proposed algorithm to find the shortest path in all graphs is substantially shorter than that required by similar existing algorithms.

Many problems in the modern world have a decentralized and distributed nature. Irregular cellular learning automata (ICLA) is a powerful mathematical model for decentralized problems and applications. Convergence of ICLA to a compatible point is very important because this convergence can provide efficient solutions for the problems. The local rule of ICLA can play a key role in this convergence. A local rule that simply rewards or punishes learning automata just based on the response of environment and actions of neighbors does not guarantee convergence of ICLA to a compatible point. In this paper, we present a new local rule that guarantees convergence to a compatible point. Formal proofs for the convergence are provided and results of the conducted experiments support our theoretical findings.

Abstract A proper semantic representation of textual information underlies many natural language processing tasks. In this paper, a novel semantic annotator is presented to generate conceptual features for text documents. A comprehensive conceptual network is automatically constructed with the aid of Wikipedia that has been represented as a Markov chain. Furthermore, semantic annotator gets a fragment of natural language text and initiates a random walk to generate conceptual features that represent topical semantic of the input text. The generated conceptual features are applicable to many natural language processing tasks where the input is textual information and the output is a decision based on its context. Consequently, the effectiveness of the generated features is evaluated in the task of document clustering and classification. Empirical results demonstrate that representing text using conceptual features and considering the relations between concepts can significantly improve not only the bag of words representation but also other state‐of‐the‐art approaches.

The group search optimiser (GSO) algorithm is a newly found evolutionary algorithm that is inspired by animal-searching behaviour and group living theory. The GSO algorithm follows the producer–scrounger framework that consists of producer, scrounger and ranger members. There are multiple key parameters in the GSO algorithm that directly affect the performance of the algorithm. Among these parameters, the maximum pursuit distance parameter plays an important role because it determines the step length of the producer and rangers of the GSO algorithm. In this paper, we develop a modified GSO algorithm by using the success rate model to adjust the maximum pursuit distance parameter of the algorithm. We test the proposed algorithm on a rich set of benchmark functions including 30- and 300-dimensional problems and compare the results with popular evolutionary and swarm algorithms. The experimental results demonstrate that the scanning mechanism of the proposed algorithm quickly optimises not only the 30-dimensional problems but also the high-dimensional (300D) problems.

One of the effective techniques for improving the rate of convergence in the particle swarm optimisation (PSO) is modifying the inertia weight parameter. This parameter can specify the search area of the swarm in the environment and establish a good balance between the global and local search ability of the particles. Several strategies have been already suggested and well tested for setting the inertia weight in static environments. However, in dynamic environments, the effect of this parameter on increasing the ability of PSO in tracking the changing optimum has been barely considered. In this paper, a time-varying inertia weight, called oscillating triangular inertia weight, is presented and its performance is measured on the moving peaks benchmark (MPB). Experimental results on various dynamic scenarios generated by MPB demonstrate that the proposed strategy has a better capability to adapt with the environmental changes in comparison with other techniques including constant inertia weight and linearly decreasing inertia weight.

Structural and behavioral parameters of many real networks such as social networks are unpredictable, uncertain, and have time-varying parameters, and for these reasons, deterministic graphs for modeling such networks are too restrictive to solve most of the real-network problems. It seems that stochastic graphs, in which weights associated to the vertices are random variables, might be better graph models for real-world networks. Once we use a stochastic graph as the model for a network, every feature of the graph such as path, spanning tree, clique, dominating set, and cover set should be treated as a stochastic feature. For example, choosing a stochastic graph as a graph model of an online social network and defining community structure in terms of clique, the concept of a stochastic clique may be used to study community structures’ properties or define spreading of influence according to the coverage of influential users; the concept of stochastic vertex covering may be used to study spread of influence. In this article, minimum vertex covering in stochastic graphs is first defined, and then four learning, automata-based algorithms are proposed for solving a minimum vertex-covering problem in stochastic graphs where the probability distribution functions of the weights associated with the vertices of the graph are unknown. It is shown that through a proper choice of the parameters of the proposed algorithms, one can make the probability of finding minimum vertex cover in a stochastic graph as close to unity as possible. Experimental results on synthetic stochastic graphs reveal that at a certain confidence level the proposed algorithms significantly outperform the standard sampling method in terms of the number of samples needed to be taken from the vertices of the stochastic graph.

Because of unpredictable, uncertain and time-varying nature of real networks it seems that stochastic graphs, in which weights associated to the edges are random variables, may be a better candidate as a graph model for real world networks. Once the graph model is chosen to be a stochastic graph, every feature of the graph such as path, clique, spanning tree and dominating set, to mention a few, should be treated as a stochastic feature. For example, choosing stochastic graph as the graph model of an online social network and defining community structure in terms of clique, and the associations among the individuals within the community as random variables, the concept of stochastic clique may be used to study community structure properties. In this paper maximum clique in stochastic graph is first defined and then several learning automata-based algorithms are proposed for solving maximum clique problem in stochastic graph where the probability distribution functions of the weights associated with the edges of the graph are unknown. It is shown that by a proper choice of the parameters of the proposed algorithms, one can make the probability of finding maximum clique in stochastic graph as close to unity as possible. Experimental results show that the proposed algorithms significantly reduce the number of samples needed to be taken from the edges of the stochastic graph as compared to the number of samples needed by standard sampling method at a given confidence level.

Abstract The emergence of knowledge repositories in a variety of domains provides a valuable opportunity for semantic interpretation of high dimensional datasets. Previous researches investigate the use of concept instead of word as a core semantic feature for incorporating semantic knowledge from an ontology into the representation model of documents. On the other hand, in machine learning and information retrieval, data objects are represented as a flat feature vector. The inconsistency between the structural nature of the knowledge repositories and the flat representation of features in machine learning leads researchers to neglect the structure of the knowledge base and leverage concepts as isolated semantic features, which is known as bag-of-concepts. Although, using concepts has some advantages over words, by neglecting the relation between concepts, the problem of vocabulary mismatch remains in force. In this paper, a novel semantic kernel is proposed which is capable of incorporating the relatedness between conceptual features. This kernel leverages clique theory to map data objects to a novel feature space wherein complex data objects will be comparable. The proposed kernel is relevant to all applications which have a prior knowledge about the relatedness between features. We concentrate on representing text documents and words using Wikipedia and WordNet, respectively. The experimental results over a set of benchmark datasets have revealed that the proposed kernel significantly improves the representation of both words and texts in the application of semantic relatedness.

Cellular learning automaton (CLA) is a recently introduced model that combines cellular automaton (CA) and learning automaton (LA). The basic idea of CLA is to use LA to adjust the state transition probability of stochastic CA. This model has been used to solve problems in areas such as channel assignment in cellular networks, call admission control, image processing, and very large scale integration placement. In this paper, an extension of CLA called irregular CLA (ICLA) is introduced. This extension is obtained by removing the structure regularity assumption in CLA. Irregularity in the structure of ICLA is needed in some applications, such as computer networks, web mining, and grid computing. The concept of expediency has been introduced for ICLA and then, conditions under which an ICLA becomes expedient are analytically found.

In this paper, an adaptive Petri net (PN), capable of adaptation to environmental changes, is introduced by the fusion of learning automata and PN. In this new model, called learning automata-based adaptive PN (APN-LA), learning automata are used to resolve the conflicts among the transitions. In the proposed APN-LA model, transitions are portioned into several sets of conflicting transitions and each set of conflicting transitions is equipped with a learning automaton which is responsible for controlling the conflicts among transitions in the corresponding transition set. We also generalize the proposed APN-LA to ASPN-LA which is a fusion between LA and stochastic PN (SPN). An application of the proposed ASPN-LA to priority assignment in queuing systems with unknown parameters is also presented.

User satisfaction as a significant antecedent to user loyalty has been highlighted by many researchers in market based literatures. SLA violation as an important factor can decrease users' satisfaction level. The amount of this decrease depends on user's characteristics. Some of these characteristics are related to QoS requirements and announced to service provider through SLAs. But some of them are unknown for service provider and selfish users are not interested to reveal them truly. Most the works in literature ignore considering such characteristics and treat users just based on SLA parameters. So, two users with different characteristics but similar SLAs have equal importance for the service provider. In this paper, we use two user's hidden characteristics, named willingness to pay for service and willingness to pay for certainty, to present a new proactive resource allocation approach with aim of decreasing impact of SLA violations. New methods based on learning automaton for estimation of these characteristics are provided as well. To validate our approach we conducted some numerical simulations in critical situations. The results confirm that our approach has ability to improve users' satisfaction level that cause to gain in profitability.

Many real world optimization problems are dynamic in which the fitness landscape is time dependent and the optima change over time such as dynamic economic modeling, dynamic resource scheduling, and dynamic vehicle routing. Such problems challenge traditional optimization methods as well as conventional evolutionary optimization algorithms. For such environments, optimization algorithms not only have to find the global optimum but also closely track its trajectory. In this paper, we propose a collaborative version of cellular PSO, named Two Phased cellular PSO to address dynamic optimization problems. The proposed algorithm introduces two search phases in order to create a more efficient balance between exploration and exploitation in cellular PSO. The conventional PSO in cellular PSO is replaced by a proposed PSO to increase the exploration capability and an exploitation phase is added to increase exploitation is the promising cells. Moreover, the cell capacity threshold which is a key parameter of cellular PSO is eliminated due to these modifications. To demonstrate the performance and robustness of the proposed algorithm, it is evaluated in various dynamic environment modeled by Moving Peaks Benchmark. The results show that for all the experimented dynamic environments, TP-CPSO outperforms all compared algorithms including cellular PSO.

Artificial fish swarm algorithm is one of the swarm intelligence algorithms which performs based on population and stochastic search contributed to solve optimization problems. This algorithm has been applied in various applications e.g. data clustering, neural networks learning, nonlinear function optimization, etc. Several problems in real world are dynamic and uncertain, which could not be solved in a similar manner of static problems. In this paper, for the first time, a modified artificial fish swarm algorithm is proposed in consideration of dynamic environments optimization. The results of the proposed approach were evaluated using moving peak benchmarks, which are known as the best metric for evaluating dynamic environments, and also were compared with results of several state-of-the-art approaches. The experimental results show that the performance of the proposed method outperforms that of other algorithms in this domain.

An efficient exploration of unknown environments is a fundamental problem in mobile robots. This paper proposes a new exploration method, in this method each specific area in environment is considered as a cell that these cells are represented by 3 abstract layers. The value of each cell in first layer is calculated by range finder's free beams. In other layers, the value of each cell is calculated by visual information, the information is received by other sensors' data and image processing that used in potential filed algorithm. We merge the value of these layers to have a single meaning value. We can use this value in many purposes e.g. finding optimal path for exploration or using this value as reward for learning methods. Then it mixed with a new improved version of A* algorithm that introduces for first time to find optimal path in unknown areas. This method implemented in official simulator of Virtual Robots League in Robocop competitions and compared with random search method. The simulation result of this method covers more unknown area compared to last methods.

Recommendation systems aim at directing users toward the resources that best meet their needs and interests. In this paper, we propose a new recommendation algorithm based on a hybrid method of distributed learning automata and graph partitioning. The proposed method utilizes usage data and hyperlink graph of the web site. The idea of the proposed method is that an appropriate recommendation for a user can be pages similar to the pages the user has already visited. To calculate similarities between pages, it is assumed that if different users request a couple of pages together, these pages are likely to correspond to the same information need therefore can be considered similar. Experiments on synthetic and real data show that the proposed algorithm provides better recommendations than the only learning automata based recommendation method reported in the literature.

Traffic signal controllers and route guidance systems are two major subsystems of Intelligent Transportation Systems that influence each other directly. Due to correlation between these subsystems, some research has been done to combine them into an integrated system. In this paper, PersianGulf - an autonomous combined traffic signal controller and route guidance system - will be proposed which has two unique features as compared to similar research. First, it is totally distributed because calculation of optimal routes are done independently in traffic signal controllers of intersections and second, it is totally autonomous because there is no need to communicate/cooperate with either traffic supervisors such as traffic management centers or the vehicles/drivers. Also, Simulation results show that PersianGulf improve the average speed of vehicles significantly in the simulated scenarios.

This paper proposes an energy-aware location-based routing protocol for mobile sensor networks that consist of frequently moving sensors. Our proposed protocol uses learning automata to select best routes that maximize delivery ratio and network lifetime. The protocol uses the location and remaining energy information of sensors to assign a cost function to each sensor node. Each node in network is equipped with a learning automaton which selects least-cost paths for each packet. Simulation results show that the proposed method achieves higher delivery ratio, lower routing overhead and lower energy consumption.

A weakly connected dominating set (WCDS) of graph G is a subset of G so that the vertex set of the given subset and all vertices with at least one endpoint in the subset induce a connected sub-graph of G. The minimum WCDS (MWCDS) problem is known to be NP-hard, and several approximation algorithms have been proposed for solving MWCDS in deterministic graphs. However, to the best of our knowledge no work has been done on finding the WCDS in stochastic graphs. In this paper, a definition of the MWCDS problem in a stochastic graph is first presented and then several learning automata-based algorithms are proposed for solving the stochastic MWCDS problem where the probability distribution function of the weight associated with the graph vertices is unknown. The proposed algorithms significantly reduce the number of samples needs to be taken from the vertices of the stochastic graph. It is shown that by a proper choice of the parameters of the proposed algorithms, the probability of finding the MWCDS is as close to unity as possible. Experimental results show the major superiority of the proposed algorithms over the standard sampling method in terms of the sampling rate.

The cellular learning automaton (CLA), which is a combination of cellular automaton (CA) and learning automaton (LA), is introduced recently. This model is superior to CA because of its ability to learn and is also superior to single LA because it is a collection of LAs which can interact with each other. The basic idea of CLA is to use LA to adjust the state transition probability of stochastic CA. Recently, various types of CLA such as synchronous, asynchronous, and open CLAs have been introduced. In some applications such as cellular networks, we need to have a model of CLA for which multiple LAs reside in each cell. In this paper, we study a CLA model for which each cell has several LAs. It is shown that, for a class of rules called commutative rules, the CLA model converges to a stable and compatible configuration. Two applications of this new model such as channel assignment in cellular mobile networks and function optimization are also given. For both applications, it has been shown through computer simulations that CLA-based solutions produce better results.

Different efforts have been made to address the problem of information overload on the Internet. Web recommendation systems based on web usage mining try to mine users' behavior patterns from web access logs, and recommend pages to the online user by matching the user's browsing behavior with the mined historical behavior patterns. In this paper we propose effective and scalable technique to solve the Web page recommendation problem. We use distributed learning automata to learn the behavior of previous users' and cluster pages based on learned pattern. One of the challenging problems in recommendation systems is dealing with unvisited or newly added pages. As they would never be recommended, we need to provide an opportunity for these rarely visited or newly added pages to be included in the recommendation set. By considering this problem, and introducing a novel Weighted Association Rule mining algorithm, we present an algorithm for recommendation purpose. We employ the HITS algorithm to extend the recommendation set. We evaluate proposed algorithm under different settings and show how this method can improve the overall quality of web recommendations.

The main goal of web pages ranking is to find the interrelated pages. In this paper, we introduce an algorithm called FPR-DLA. In the proposed method learning automata is assigned to each web page which its function is determining the weight of hyperlinks between web pages. Also for determining the weight of each web page parameters such as time duration on a web page and the importance of web pages are considered. Time duration on a web page and the importance of web pages are characterized as a fuzzy linguistic variable. The proposed algorithm calculates the rank of each web page as recursive according to the weights of each web page and hyperlinks between web pages. Experimental results show that the proposed method has a considerable efficiency in determining the rank of web pages.

A solution to channel assignment problem in cellular networks is self-organizing channel assignment algorithm with distributed control. In this paper, we propose three cellular learning automata based dynamic channel assignment algorithms. In the first two algorithms, no information about the status of channels in the whole network will be used by cells for channel assignment whereas in the third algorithm, the additional information regarding status of channels may be gathered and then used by cells in order to allocate channels. The simulation results show that by using the proposed channel assignment algorithms the micro-cellular network can self-organize itself. The simulation results also show that the additional information used by the third algorithm help the cellular learning automata to find an assignment which results in lower blocking probability of calls for the network.

In this paper, two learning automata based adaptive limited fractional guard channel algorithms for cellular mobile networks are proposed. These algorithms try to minimize the blocking probability of new calls subject to the constraint on the dropping probability of the handoff calls. To evaluate the proposed algorithms, computer simulations are conducted. The simulation results show that the performance of the proposed algorithms are close to the performance of the limited fractional guard channel algorithm for which prior knowledge about traffic parameters are needed. The simulation results also show that the proposed algorithms outperforms the recently introduced dynamic guard channel algorithms.

Clustering is currently one of the most crucial techniques for dealing with massive amount of heterogeneous information on the web, which is beyond human beingpsilas capacity to digest. Recent studies have shown that the most commonly used partitioning-based clustering algorithm, the K-means algorithm, is more suitable for large datasets. However, the K-means algorithm can generate a local optimal solution. In this paper we present novel harmony search clustering algorithms that deal with documents clustering based on harmony search optimization method. By modeling clustering as an optimization problem, first, we propose a pure harmony search based clustering algorithm that finds near global optimal clusters within a reasonable time. Contrary to the localized searching of the K-means algorithm, the harmony search clustering algorithm performs a globalized search in the entire solution space. Then harmony clustering is integrated with the K-means algorithm in three ways to achieve better clustering. The proposed algorithms improve the K-means algorithm by making it less dependent on the initial parameters such as randomly chosen initial cluster centers, hence more stable. In the experiments we conducted, we applied the proposed algorithms, K-means clustering algorithm on five different document datasets. Experimental results reveal that the proposed algorithms can find better clusters when compared to K-means and the quality of clusters is comparable and converge to the best known optimum faster than it.

Cellular learning automata (CLA) which is obtained by combining cellular automata (CA) and learning automata (LA) models is a mathematical model for dynamical complex systems that consists of a large number of simple learning components. CLA- EC, introduced recently is an evolutionary algorithm which is obtained by combining CLA and evolutionary computation (EC). In this paper CLA-EC with recombination operator is introduced. Recombination increases explorative behavior of CLA-EC and also provides a mechanism for partial structure exchange between chromosomes of population individuals that standard CLA-EC is not capable of performing it. This modification greatly improves CLA-EC ability to effectively search solution space and leave local optima. Experimental results on five optimization test functions show the superiority of this new version of CLA-EC over the standard CLA-EC.

In recent years, global and grid computing emerge as two powerful technology trends. In this paper, we compare these two approaches of distributed computing. First, we present a definition for global computing that accentuates the key point in this trend. This key point distinguishes global computing from other trends and covers many such systems. Second, we contrast two approaches in general characteristics. Then, by comparing them in technical issues, we show that the key point in our definition of global computing is the main source of many technical differences between these methods. Finally, we present our opinion about the probable future of global and grid computing

In this paper, we first introduce a network of learning automata, which we call it as distributed learning automata and then propose some iterative algorithms for solving stochastic shortest path problem. These algorithms use distributed learning automata to find a policy that determines a path from a source node to a destination node with minimal expected cost (length). In these algorithms, at each stage distributed learning automata determines which edges to be sampled. This sampling method may result in decreasing unnecessary samples and hence decreasing the running time of algorithms. It is shown that the shortest path is found with a probability as close as to unity by proper choice of the parameters of the proposed algorithms.

In this paper, we study the dynamical properties of the population based incremental learning (PBIL) algorithm when it uses truncation, proportional, and Boltzmann selection schemas. The results show that if the population size tends to infinity, with any learning rate, the local optima of the function to be optimized are asymptotically stable fixed points of the PBIL

In this paper we introduce an estimation of distribution algorithm based on a team of learning automata. The proposed algorithm is a model based search optimization method that uses a team of learning automata as a probabilistic model of high quality solutions seen in the search process. Simulation results show that the proposed algorithm is a good candidate for solving optimization problems.

The CLA-EC is a model obtained by combining the concepts of cellular learning automata and evolutionary algorithms. The parallel structure of the CLA-EC makes it suitable for hardware-based applications including evolvable hardware. In this paper, based on the SIMD model, a parallel architecture is proposed and implemented on FPGA. Simulation results show that the proposed architecture can solve optimization problems thousands times faster than the sequential implementations.

Case-based reasoning (CBR) is a knowledge-based problem-solving technique, which is based on reuse of previous experiences. We propose a new model for static task assignment in heterogeneous computing systems. The proposed model is a combination of the case based reasoning and the learning automata model. In this new model a learning automata model is used as adaptation mechanism, which adapts previously experienced cases to the problem to be solved. The objective of the proposed model is to reduce the number of iterations required to find a semioptimum solution. The application is modeled as a set of independent tasks and the heterogeneous computing system is modeled as a network of machines. Using computer simulation, it is shown that the combined model outperforms the model that only uses learning automata.

In this paper, a new fuzzy clustering algorithm that uses cellular learning automata based evolutionary computing (CLA-EC) is proposed. The CLA-EC is a model obtained by combining the concepts of cellular learning automata and evolutionary algorithms. The CLA-EC is used to search for cluster centers in such a way that minimizes the clustering criterion. The simulation results indicate that the proposed algorithm produces clusters with acceptable quality with respect to clustering criterion and provides a performance that is superior to that of the C-means algorithm.

The particle swarm is one of the most powerful methods for solving global optimization problems. This method is an adaptive algorithm based on social-psychological metaphor. A population of particle adapts by returning stochastically toward previously successful regions in the search space and is influenced by the successes of their topological neighbors. In this paper we propose a learning automata based discrete binary particle swarm algorithm. In the proposed algorithm the set of learning automata assigned to a particle may be viewed as the brain of the particle determining its position from its own and other particles past experience. Simulation results show that the proposed algorithm is a good candidate for solving optimization problems.

In this paper we propose a new fuzzy-based navigation system for two intelligent mobile robots using distributed value function reinforcement learning. The robots use their sensors to provide information about their workspace. A fuzzy controller uses this information to select a proper action for the currently sensed state. The parameters of the input and output fuzzy membership functions are determined by a learning automation at each time step based on the sparseness of the obstacles. Therefore the robots learn to control their velocity, and attention range regarding the density of the obstacles in the workspace. The distributed approach enables the robots to learn more than one simple behavior concurrently. So, in contrast to the existing methods no behavior blending is needed. This approach also enables the robots to learn a value function, which is an estimate of future rewards for both of them. In other words cooperation is maintained and each robot learns to execute the actions that are good for both of them. The proposed controller has a very simple architecture and clear logic. The time and computation cost are low and it can adapt well to environment changes. Computer simulations are used to investigate the effectiveness of the controller.

One popular learning algorithm for feedforward neural networks is the backpropagation (BP) algorithm which includes parameters, learning rate (eta), momentum factor (alpha) and steepness parameter (lambda). The appropriate selections of these parameters have large effects on the convergence of the algorithm. Many techniques that adaptively adjust these parameters have been developed to increase speed of convergence. In this paper, we shall present several classes of learning automata based solutions to the problem of adaptation of BP algorithm parameters. By interconnection of learning automata to the feedforward neural networks, we use learning automata scheme for adjusting the parameters eta, alpha, and lambda based on the observation of random response of the neural networks. One of the important aspects of the proposed schemes is its ability to escape from local minima with high possibility during the training period. The feasibility of proposed methods is shown through simulations on several problems.

Despite of the many successful applications of backpropagation for training multi-layer neural networks, it has many drawbocks. For complex problems it may require a long time to train the networks, and it may not train at all. Long training time can be the result of the non-optimal parameters. It is not easy to choose appropriate value of the parameters for a particular problem. In this paper, by interconnection of fixed structure learning automata (FSLA) to the feedforward neural networks, we apply learning automata (LA) scheme for adjusting these parameters based on the observation of random response of neural networks. The main motivation in using learning automata as an adaptation algorithm is to use its capability of global optimization when dealing with multi-modal surface. The feasibility of proposed method is shown through simulations on three learning problems: exclusive-or, encoding problem, and digit recognition. The simulation results show that the adaptation of these parameters using this method not only increases the convergence rate of learning but it increases the likelihood of escaping from the local minima.