基于遗传算法和强化学习的贝叶斯网络结构学习算法

遗传算法是基于自然界中生物遗传规律的适应性原则对问题解空间进行搜寻和最优化的方法。贝叶斯网络是对不确定性知识进行建模、推理的主要方法,Bayesian网中的学习问题(参数学习与结构学习)是个NP-hard问题。强化学习是利用新顺序数据来更新学习结果的在线学习方法。介绍了利用强化学习指导遗传算法,实现对贝叶斯网结构进行有效学习。     

基于遗传算法的本体概念分类规则学习算法

为了更好地界定本体中的概念,提出一种基于遗传算法（Genetic Algorithm,GA）的本体概念分类规则的学习方法．从已有的本体库中获取实例作为训练样本,通过该算法寻找一组与数据样本集一致的规则．以一组规则集作为遗传算法的个体,即优化的目标,同时考虑到规则集的覆盖性、一致性、简洁性和多样性4个方面建立适应值函数,优化得到一组能够分类概念的规则集合．进而这组规则集可用于指导和丰富本体知识,例如当本体中引入新的实例时,可以通过此概念分类规则集确定实例所属的概念．对已有本体学习后的实验结果表明该算法收敛性很好,而且能获得较好的规则集．     

Annealing-based heuristics and genetic algorithms for circuit partitioning in parallel test generation

In this paper simulated annealing and genetic algorithms are applied to the graph partitioning problem. These techniques mimic processes in statistical mechanics and biology, respectively, and are the most popular meta-heuristics or general-purpose optimization strategies. A hybrid algorithm for circuit partitioning, which uses tabu search to improve the simulated annealing meta-heuristics, is also proposed and compared with pure tabu search and simulated annealing algorithms, and also with a genetic algorithm. The solutions obtained are compared and evaluated by including the hybrid partitioning algorithm in a parallel test generator which is used to determine the test patterns for the circuits of the frequently used ISCAS benchmark set.     

On-line free-flight path optimization based on improved genetic algorithms

Free flight (FF) is the ideal strategy of current investigations on air traffic management systems, where an on-line flight path optimization algorithm is of top importance. This paper proposes an innovative algorithm with potential real-time properties for FF path optimization, by using an improved genetic algorithm (GA). Two kinds of mathematical models for the on-line flight path optimization problem are proposed to cover the near and far future applications. Several improvements are introduced to the GA to speed up its convergence as well as to improve performance. Simulation results show that the new algorithm is effective and has potential to solve the on-line FF path optimization problem in real time.     

基于选择性集成遗传算法的BNC结构学习

为克服IQ算法在处理贝叶斯网络分类器（Bayesian Network Classifier,BNC）结构学习中要求先指定适合节点次序的缺点,提出GA-K2算法,将基于选择性集成的整数编码遗传算法引入到K2算法中,使之能得到最佳节点次序并且网络结构收敛到全局最优．构建贝叶斯网络分类器进行分类,实验结果表明GA-K2算法优于随意指定节点顺序的IQ算法．     

交互式遗传算法基于用户认知不确定性的定向变异

利用用户认知的不确定性设计定向变异算子．首先,采用主成分分析法辨识用户认知的不确定性;然后,给出用户认知不确定性的区间表示与更新策略;最后,将用户认知的不确定性指导定向变异算子,包括：选择待变异的进化个体,确定变异位置,以及变异方法等．将所提方法应用于人眼图形优化,实验结果验证了该方法的优越性．     

CPGEA: a grouping genetic algorithm for material cutting plan generation

Construction firms specializing in large commercial buildings often purchase large steel plates, cut them into pieces and then weld the pieces into H-beams and other construction components. We formalize the material ordering and cutting problem faced by this industry and propose a grouping genetic algorithm, called CPG_EA, for efficiently controlling the relevant costs. We test the quality of CPG_EA in various ways. Three sets of simulated problems with known optimal solutions are solved using CPG_EA, and the gap between its solutions and optimal solutions is measured. The same problem sets are also solved with an expert system and a multi-start greedy heuristic. CPG_EA solutions are found to be consistently lower cost than the competing methods. The difference in solution quality is most pronounced for difficult problems requiring multiple identical plates in the optimal solution. CPG_EA is also tested using data from actual construction projects of a company faced with this problem. Since an optimal solution for the problems is not available, a lower bound is created. For the historical problems tested, the average percent difference between CPG_EA solutions and the lower bound is 0.67%. To put this performance in context, the results of solving these problems with an expert system and using experienced engineers is also reported. Of these three methods, CPG_EA achieves the best performance and the human experts the worst performance.     

An impulse control approach to spacecraft autonomous rendezvous based on genetic algorithms

This paper proposes a novel approach to spacecraft impulse autonomous rendezvous by using genetic algorithms. Based on the Clohessy-Wiltshire (C-W) equations, the whole rendezvous process is described as a switching system composed of closed-loop system and open-loop system, which correspond to the impulse action phase and free motion phase during the rendezvous process. Based on Lyapunov theory, the autonomous rendezvous problem is regarded as an asymptotic stabilization problem. By introducing two virtual energy functions, the stability of the switching system is analyzed, and the duration of the impulse action and the thrust limitation are considered synthetically. Then, a state-feedback controller design method is proposed, and an approach based on linear matrix inequality and genetic algorithm (GA) is proposed to solve the controller design problem and the calculation steps are presented. With the designed controller, the impulse thrust which satisfies the given thrust constraint is determined according to the real-time relative state between two spacecraft at the impulse instant, and the impulse duration is kept as short as possible. The effectiveness of the proposed approach is illustrated by simulation examples.     

Leveraging saving-based algorithms by master-slave genetic algorithms

Saving-based algorithms are commonly used as inner mechanisms of efficient heuristic construction procedures. We present a general mechanism for enhancing the effectiveness of such heuristics based on a two-level genetic algorithm. The higher-level algorithm searches in the space of possible merge lists which are then used by the lower-level saving-based algorithm to build the solution. We describe the general framework and we illustrate its application to three hard combinatorial problems. Experimental results on three hard combinatorial optimization problems show that the approach is very effective and it enables considerable enhancement of the performance of saving-based algorithms.     

Multi-operator based evolutionary algorithms for solving constrained optimization problems

Over the last two decades, many sophisticated evolutionary algorithms have been introduced for solving constrained optimization problems. Due to the variability of characteristics in different COPs, no single algorithm performs consistently over a range of problems. In this paper, for a better coverage of the problem characteristics, we introduce an algorithm framework that uses multiple search operators in each generation. The appropriate mix of the search operators, for any given problem, is determined adaptively. The framework is tested by implementing two different algorithms. The performance of the algorithms is judged by solving 60 test instances taken from two constrained optimization benchmark sets from specialized literature. The first algorithm, which is a multi-operator based genetic algorithm (GA), shows a significant improvement over different versions of GA (each with a single one of these operators). The second algorithm, using differential evolution (DE), also confirms the benefit of the multi-operator algorithm by providing better and consistent solutions. The overall results demonstrated that both GA and DE based algorithms show competitive, if not better, performance as compared to the state of the art algorithms.     

Semantic relation based personalized ranking approach for engineering document retrieval

Since engineering design is heavily informational, engineers want to retrieve existing engineering documents accurately during the product development process. However, engineers have difficulties searching for documents because of low retrieval accuracy. One of the reasons for this is the limitation of existing document ranking approaches, in which relationships between terms in documents are not considered to assess the relevance of the retrieved documents. Therefore, we propose a new ranking approach that provides more correct evaluation of document relevance to a given query. Our approach exploits domain ontology to consider relationships among terms in the relevance scoring process. Based on domain ontology, the semantics of a document are represented by a graph (called Document Semantic Network) and, then, proposed relation-based weighting schemes are used to evaluate the graph to calculate the document relevance score. In our ranking approach, user interests and searching intent are also considered in order to provide personalized services. The experimental results show that the proposed approach outperforms existing ranking approaches. A precisely represented semantics of a document as a graph and multiple relation-based weighting schemes are important factors underlying the notable improvement.     

Personalized mobile English vocabulary learning system based on item response theory and learning memory cycle

Since learning English is very popular in non-English speaking countries, developing modern assisted-learning tools that support effective English learning is a critical issue in the English-language education field. Learning English involves memorization and practice of a large number of vocabulary words and numerous grammatical structures. Vocabulary learning is a principal issue for English learning because vocabulary comprises the basic building blocks of English sentences. Therefore, many studies have attempted to improve the efficiency and performance when learning English vocabulary. With the accelerated growth in wireless and mobile technologies, mobile learning using mobile devices such as PDAs, tablet PCs, and cell phones has gradually become considered effective because it inherits all the advantages of e-learning and overcomes limitations of learning time and space that limit web-based learning systems. Therefore, this study presents a personalized mobile English vocabulary learning system based on Item Response Theory and learning memory cycle, which recommends appropriate English vocabulary for learning according to individual learner vocabulary ability and memory cycle. The proposed system has been successfully implemented on personal digital assistant (PDA) for personalized English vocabulary learning. The experimental results indicated that the proposed system could obviously promote the learning performances and interests of learners due to effective and flexible learning mode for English vocabulary learning.     

Orderly algorithms for graph generation

The existence of orderly analogues of graph generators proposed by Heap and Farrell is established. The modifications to these algorithms supply practical methods enabling one to generate exhaustive lists of graphs and locally restricted graphs; moreover, the difficulty involved in ensuring that no duplications occur in the list is greatly reduced.     

基于改进遗传算法的最小生成树算法

以图论和改进遗传算法为基础,提出了一种求最小生成树的遗传算法。该算法采用二进制表示最小树问题,并设计出相应的适应度函数、算子以及几种控制策略,以提高执行速度和进化效率。传统算法一次只能得到一个候选解。用该算法对其求解,可以在较短的时间内以较高的概率获得多个候选解。应用实例表明该算法优于传统算法。     

Automatic generation of basis test paths using variable length genetic algorithm

Path testing is the strongest coverage criterion in white box testing. Finding target paths is a key challenge in path testing. Genetic algorithms have been successfully used in many software testing activities such as generating test data, selecting test cases and test cases prioritization. In this paper, we introduce a new genetic algorithm for generating test paths. In this algorithm the length of the chromosome varies from iteration to another according to the change in the length of the path. Based on the proposed algorithm, we present a new technique for automatically generating a set of basis test paths which can be used as testing paths in any path testing method. The proposed technique uses a method to verify the independency of the generated paths to be included in the basis set of paths. In addition, this technique employs a method for checking the feasibility of the generated paths. We introduce new definitions for the key concepts of genetic algorithm such as chromosome representation, crossover, mutation, and fitness function to be compatible with path generation. In addition, we present a case study to show the efficiency of our technique. We conducted a set of experiments to evaluate the effectiveness of the proposed path generation technique. The results showed that the proposed technique causes substantial reduction in path generation effort, and that the proposed GA algorithm is effective in test path generation.     

Two-dimensional packing problems using genetic algorithms

This paper presents a technique for applying genetic algorithms for the two-dimensional packing problem. The approach is applicable to not only convex shaped objects, but can also accommodate any type of concave and complex shaped objects including objects with holes. In this approach, a new concept of a two-dimensional genetic chromosome is introduced. The total layout space is divided into a finite number of cells for mapping it into this 2D genetic algorithm chromosome. The mutation and crossover operators have been modified and are applied in conjunction with connectivity analysis for the objects to reduce the creation of faulty generations. A new feature has been added to the Genetic Algorithm (GA) in the form of a new operator called compaction. Several examples of GA-based layout are presented.     

基于改进遗传算法的本体概念分类规则算法

针对本体学习中的概念学习,提出了一种基于改进遗传算法的本体概念规则的学习方法。该方法在传统遗传算法的遗传操作算予中引入“杂交优势”思想对交叉算子进行了改进,并加强了变异算子的对算法的影响;同时算法在执行过程中对训练样本集使用了约减策略,从而找出了一个能正确覆盖样本空间中所有实例并且不覆盖任何错误实例的规则集合。     

Genetic algorithm based multi-agent system applied to test generation

Automatic test generating system in distributed computing context is one of the most important links in on-line evaluation system. Although the issue has been argued long since, there is not a perfect solution to it so far. This paper proposed an innovative approach to successfully addressing such issue by the seamless integration of genetic algorithm (GA) and multi-agent system. In the design phase, a test ontology was firstly defined for smoothing the communication among agents. For the implementation of GA, The fitness function and the structure of chromosome were identified on the basis of the analysis of constraint conditions associated with a test. To demonstrate the task execution flow and messages passing among agents, the activity diagram and sequence diagram were also shown on the AUML basis. In the phase of implementation, the JADE based agent behavior model was described in detail and the implementation platform was also demonstrated. The final simulation results validated the feasibility of the proposed approach.     

基于遗传算法的人工神经网络的应用综述

本文介绍了一种新型的、随机性的全局优化方法－遗传算法,系统地讨论了基于遗传算法的人工神经网络的主要应用,并给出了大量实验数据,实验结果表明遗传算法具有快速学习网络权重的能力,并且能够摆脱局部极点的困扰。     

基于WUM的个性化智能站点的研究

Internet和电子商务的发展带动了Web数据挖掘技术的发展.本文提出了针对提高Web服务质量的解决方案:采用WUM为核心技术,建立一个智能Web站点,通过建立与更新用户模式库,个性化地为用户的访问提供推荐服务.