Aurora T. R. Pozo

A symbolic fault-prediction model based on multiobjective particle swarm optimization

In the literature the fault-proneness of classes or methods has been used to devise strategies for reducing testing costs and efforts. In general, fault-proneness is predicted through a set of design metrics and, most recently, by using Machine Learning (ML) techniques. However, some ML techniques cannot deal with unbalanced data, characteristic very common of the fault datasets and, their produced results are not easily interpreted by most programmers and testers. Considering these facts, this paper introduces a novel fault-prediction approach based on Multiobjective Particle Swarm Optimization (MOPSO). Exploring Pareto dominance concepts, the approach generates a model composed by rules with specific properties. These rules can be used as an unordered classifier, and because of this, they are more intuitive and comprehensible. Two experiments were accomplished, considering, respectively, fault-proneness of classes and methods. The results show interesting relationships between the studied metrics and fault prediction. In addition to this, the performance of the introduced MOPSO approach is compared with other ML algorithms by using several measures including the area under the ROC curve, which is a relevant criterion to deal with unbalanced data.

Modeling software reliability growth with genetic programming

Reliability models are very useful to estimate the probability of the software fail along the time. Several different models have been proposed to estimate the reliability growth, however, none of them has proven to perform well considering different project characteristics. In this work, we explore genetic programming (GP) as an alternative approach to derive these models. GP is a powerful machine learning technique based on the idea of genetic algorithms and has been acknowledged as a very suitable technique for regression problems. The main motivation to choose GP for this task is its capability of learning from historical data, discovering an equation with different variables and operators. In this paper, experiments were conducted to confirm this hypotheses and the results were compared with traditional and neural network models

A Genetic Programming Approach for Software Reliability Modeling

Genetic programming (GP) models adapt better to the reliability curve when compared with other traditional, and non-parametric models. In a previous work, we conducted experiments with models based on time, and on coverage. We introduced an approach, named genetic programming and Boosting (GPB), that uses boosting techniques to improve the performance of GP. This approach presented better results than classical GP, but required ten times the number of executions. Therefore, we introduce in this paper a new GP based approach, named (¿ + ¿) GP. To evaluate this new approach, we repeated the same experiments conducted before. The results obtained show that the (¿ + ¿) GP approach presents the same cost of classical GP, and that there is no significant difference in the performance when compared with the GPB approach. Hence, it is an excellent, less expensive technique to model software reliability.

Exploring Genetic Programming and Boosting Techniques to Model Software Reliability

Software reliability models are used to estimate the probability that a software fails at a given time. They are fundamental to plan test activities, and to ensure the quality of the software being developed. Each project has a different reliability growth behavior, and although several different models have been proposed to estimate the reliability growth, none has proven to perform well considering different project characteristics. Because of this, some authors have introduced the use of Machine Learning techniques, such as neural networks, to obtain software reliability models. Neural network-based models, however, are not easily interpreted, and other techniques could be explored. In this paper, we explore an approach based on genetic programming, and also propose the use of boosting techniques to improve performance. We conduct experiments with reliability models based on time, and on test coverage. The obtained results show some advantages of the introduced approach. The models adapt better to the reliability curve, and can be used in projects with different characteristics.

Bayesian automatic programming

In this work a new approach, named Bayesian Automatic Programming (BAP), to inducing programs is presented. BAP integrates the power of grammar evolution and probabilistic models to evolve programs. We explore the use of BAP in two domains: a regression problem and the artificial ant problem. Its results are compared with traditional Genetic Programming (GP). The experimental results found encourage further investigation, especially to explore BAP in other domains and to improve the proposed approach to incorporating new mechanisms.

A multi-objective optimization approach for the integration and test order problem

A common problem found during the integration testing is to determine an order to integrate and test the units. Important factors related to stubbing costs and constraints regarding to the software development context must be considered. To solve this problem, the most promising results were obtained with multi-objective algorithms, however few algorithms and contexts have been addressed by existing works. Considering such fact, this paper aims at introducing a generic approach based on multi-objective optimization to be applied in different development contexts and with distinct multi-objective algorithms. The approach is instantiated in the object and aspect-oriented contexts, and evaluated with real systems and three algorithms: NSGA-II, SPEA2 and PAES. The algorithms are compared by using different number of objectives and four quality indicators. Results point out that the characteristics of the systems, the instantiation context and the number of objectives influence on the behavior of the algorithms. Although for more complex systems, PAES reaches better results, NSGA-II is more suitable to solve the referred problem in general cases, considering all systems and indicators.

Establishing integration test orders of classes with several coupling measures

During the inter-class test, a common problem, named Class Integration and Test Order (CITO) problem, involves the determination of a test class order that minimizes stub creation effort, and consequently test costs. The approach based on Multi-Objective Evolutionary Algorithms (MOEAs) has achieved promising results because it allows the use of different factors and measures that can affect the stubbing process. Many times these factors are in conflict and usually there is no a single solution for the problem. Existing works on MOEAs present some limitations. The approach was evaluated with only two coupling measures, based on the number of attributes and methods of the stubs to be created. Other MOEAs can be explored and also other coupling measures. Considering this fact, this paper investigates the performance of two evolutionary algorithms: NSGA-II and SPEA2, for the CITO problem with four coupling measures (objectives) related to: attributes, methods, number of distinct return types and distinct parameter types. An experimental study was performed with four real systems developed in Java. The obtained results point out that the MOEAs can be efficiently used to solve this problem with several objectives, achieving solutions with balanced compromise between the measures, and of minimal effort to test.

Effective Linkage Learning Using Low-Order Statistics and Clustering

The adoption of probabilistic models for selected individuals is a powerful approach for evolutionary computation. Probabilistic models based on high-order statistics have been used by estimation of distribution algorithms (EDAs), resulting better effectiveness when searching for global optima for hard optimization problems. This paper proposes a new framework for evolutionary algorithms, which combines a simple EDA based on order 1 statistics and a clustering technique in order to avoid the high computational cost required by higher order EDAs. The algorithm uses clustering to group genotypically similar solutions, relying that different clusters focus on different substructures and the combination of information from different clusters effectively combines substructures. The combination mechanism uses an information gain measure when deciding which cluster is more informative for any given gene position, during a pairwise cluster combination. Empirical evaluations effectively cover a comprehensive range of benchmark optimization problems.

Controversy Corner: Search Based Software Engineering: Review and analysis of the field in Brazil

Search Based Software Engineering (SBSE) is the field of software engineering research and practice that applies search based techniques to solve different optimization problems from diverse software engineering areas. SBSE approaches allow software engineers to automatically obtain solutions for complex and labor-intensive tasks, contributing to reduce efforts and costs associated to the software development. The SBSE field is growing rapidly in Brazil. The number of published works and research groups has significantly increased in the last three years and a Brazilian SBSE community is emerging. This is mainly due to the Brazilian Workshop on Search Based Software Engineering (WOES), co-located with the Brazilian Symposium on Software Engineering (SBES). Considering these facts, this paper presents results of a mapping we have performed in order to provide an overview of the SBSE field in Brazil. The main goal is to map the Brazilian SBSE community on SBES by identifying the main researchers, focus of the published works, fora and frequency of publications. The paper also introduces SBSE concerns and discusses trends, challenges, and open research problems to this emergent area. We hope the work serves as a reference to this novel field, contributing to disseminate SBSE and to its consolidation in Brazil.

Using archiving methods to control convergence and diversity for Many-Objective Problems in Particle Swarm Optimization

Multi-Objective Particle Swarm Optimization (MOPSO) is a population based multi-objective meta-heuristic inspired on animal swarm intelligence. It is used to solve several Multi-Objective Optimization Problems (MOPs), problems with more than one objective function. However, Multi-Objective Evolutionary Algorithms (MOEAs), including MOPSO, have some limitations when the number of objective grows. Many-Objective Optimization research methods to decrease the negative effect of applying MOEAs into problems with more than three objective functions. In this context, the goal of this work is to explore several archiving methods from the literature used by MOPSO to store the selected leaders into Many-Objective Problems. Moreover, new archiving methods are proposed specially for these problems. The use of the archiving methods into MOPSO is evaluated through an empirical analysis aiming to observe the impact of these methods in the convergence and the diversity to the Pareto front, in Many-Objective scenarios.