Catálogo de publicaciones - libros

In this paper, five previous Particle Swarm Optimization (PSO) algorithms for multimodal function optimization are reviewed. A new and a successful PSO based algorithm, named as CPSO is proposed. CPSO enhances the exploration and exploitation capabilities of PSO by performing search using a random walk and a hill climbing components. Furthermore, one of the previous PSO approaches is improved incredibly by means of a minor adjustment. All algorithms are compared over a set of well-known benchmark functions.

The purpose of this paper is to generalize Quantum-behaved Particle Swarm Optimization (QPSO) Algorithm to discrete binary search space. To design Binary QPSO (BQPSO), we redefine the position vector and the distance between two positions, and adjust the iterative equations of QPSO to binary search space. The operations designed for BQPSO are far different from those in BPSO, but somewhat like those in Genetic Algorithms (GAs). Therefore, BQPSO integrates strongpoint of GA with the features of PSO, which make it able to find out the global optimum of the problem more efficiently than BPSO, as shown by the experiment results of BQPSO and BPSO on De Jong’s five test functions.

The paper presents an artificial world model in which various self-organization and self-modification processes could be simulated. The model is a two dimensional space in which there are stacks of hexagonal tiles which are moving, colliding, and making bonds between them. On the higher level of organization a structure of tiles specifies some function whose execution affects other tiles in its neighborhood. The functions encoded in the structures of tiles are expressed in the simple Prolog like language. Few examples illustrate the behavior of the system.

Based on the previous proposed Quantum-behaved Particle Swarm Optimization (QPSO), in this paper, a novel and more efficient search strategy with a selection operation is introduced into QPSO to improve the search ability of QPSO. While the center of position distribution of each particle in QPSO is determined by global best position and personal best position, in the Modified QPSO (MQPSO), the global best position is substituted by a personal best position of a randomly selected particle. The MQPSO also maintains the mean best position of the swarm as in the previous QPSO to make the swarm more efficient in global search. The experiment results on benchmark functions show that MQPSO has stronger global search ability than QPSO and PSO.

In the paper, we describe an application of stochastic context-free grammars (SCFG) to modelling of the formal RNA string language. The simplification of the stochastic context-free grammar and it’s conversion to Chomsky normal form was used. We present the modification of Cocke-Kasami-Younger algorithm that is used for probabilistic estimations of stochastic grammars for RNA sequences. Some better algorithms were constructed to decrease the computational complexity but still on the level of () where is the length of the RNA strings. The results of using the algorithms to the training sample consisted of tRNA chains of Acinetobacter sp. bactery are described.

The paper presents a new way of modeling the human performance in two player games using a rating system based on Kelly Criterion which is often utilized for gambling and financial engineering. The advantage of the proposed system is the ability to assess playing strength based on both the final outcome of the game and the style of play. The last aspect of the rating assessment system is novel compared to the rating systems developed so far (like ELO, Bradley-Terry, etc.). Another advantage of the proposed method is the tackling of the problem of drawn games. To the very best authors knowledge the approach presented below is a relatively new look at the problem of playing strength assessment based on probability theory. The paper presents and discusses few illustrative examples.

Following [4], we analyze boosting from a game-theoretic perspective. We define a wide class of boosting classification algorithms called H-boosting methods, which are based on Hannan-consistent game playing strategies. These strategies tend to minimize the regret of a player, i.e. are able to minimize the difference between its expected cumulative loss and the cumulative loss achievable using the single best strategy. The “weighted majority” boosting algorithm [4] is proved to belong to the class of H-boosting procedures. A new boosting algorithm is proposed, as an another example of such a regret-minimizing method.

A new method of genetic evolution of linear and nonlinear evaluation functions in the game of checkers is presented. Several practical issues concerning application of genetic algorithms for this task are pointed out and discussed. Experimental results confirm that proposed approach leads to efficient evaluation functions comparable to the ones used in some of commercial applications.

Recently, there are many researches about NPC (Non-player character) which are appeared in many games. There is a little study which has view of the speech of NPC, but most of them are focused on the movement of NPC. Most of NPC say the same line repeatedly, that make the game unrealistic. In the paper, we construct the quest ontology of game World of Warcraft and collect the articles of Q/A bulletin board which is from the web site, also called community, of the game. With the game ontology and the article from the bulletin, we construct a corpus consisted of game keywords as NPC knowledge. In the runtime, when a player asks a question to NPC, it can find the answer by comparing the question and its own knowledge.

The saplings sowing and growing up algorithm (SGA) was inspired by a natural events – evolution of growing up of trees. This algorithm contains two phases: Sowing Phase and Growing up Phase. In this paper, the theoretical foundations of SGA were determined. SGA is defined as a computational model, and it was depicted that there are a collection of Turing Machines for simulating SGA.