Om Prakash Sangwan

Software reliability prediction modeling: A comparison of parametric and non-parametric modeling

Reliable softwares are the need of modern digital era. Failure nonlinearity makes software reliability a complicated task. Over past decades, many researchers have contributed many parametric / non parametric software reliability growth models and discussed their assumptions, applicability and predictability. It concluded that traditional parametric software reliability models have many shortcomings related to their unrealistic assumptions, environment-dependent applicability, and questionable predictability. In contrast to parametric software reliability growth models, the non-parametric software reliability growth models which use machine learning techniques or time series modeling have been proposed by researchers. This paper evaluates and compares the accuracy of 2 parametric and 2 non parametric software reliability growth models on 3 real-life data sets for software failures.

Parameter Estimation of Software Reliability Model Using Firefly Optimization

This paper, presents an effective parameter estimation technique for software reliability growth models using firefly algorithm. Software failure rate with respect to time has always been a foremost concern in the software industry. Every second organization aims to achieve defect free software products, which makes software reliability prediction a burning research area. Software reliability prediction techniques generally use numerical estimation method for parameter estimation, which is certainly not the best. Local optimization, biasness and model’s parameter initialization are some foremost limitation, which eventually suffers the finding of optimal model parameters. Firefly optimization overcomes these limitations and provides optimal solution for parameter estimation of software reliability growth models. Goel Okumoto model and Vtub based fault detection rate model is selected to validate the results. Seven real world datasets were used to compare the proposed technique against Cuckoo search technique and CASRE tool. The results indicate the superiority of proposed approach over existing numerical estimation techniques.

An Insight of software quality models applied in predicting software quality attributes: A comparative analysis

Software is crucial in giving a competitive edge to most of the organizations. Software has become main part of business products, systems and services. Software products quality was seen as a significant element in success of business. The purpose of the research is to the existing models related to software quality which is used for predicting the quality attributes in the software. SQM models selected for this study to compare are ISO 9126, Bertoa, ISO 25010, CBQM and Alvaro model. Main characteristics selected are portability, maintainability, flexibility, usability, reliability and efficiency and sub characteristics are accuracy, testability, extendibility, compatibility, understandability and performance. This study adopts secondary source of data. It was noticed that for the selected set of quality attributes characteristics, Alvaro model excel well in main characteristics as well as other sub characteristics namely understandability, accuracy and testability.

Component-Based Quality Prediction via Component Reliability Using Optimal Fuzzy Classifier and Evolutionary Algorithm

Sequentially to meet the rising necessities, software system has become more complex for software support from profuse varied areas. In software reliability engineering, many techniques are available to ensure the reliability and quality. In design models, prediction techniques play an important role. In case of component-based software systems, accessible reliability prediction approaches experience the following drawbacks and hence restricted in their applicability and accuracy. Here, we compute the application reliability which is estimated depend upon the reliability of the individual components and their interconnection mechanisms. In our method, the quality of the software can be predicted in terms of reliability metrics. After that the component-based feature extraction, the reliability is calculated by optimal fuzzy classifier (OFC). Here, the fuzzy rules can be optimized by evolutionary algorithms. The implementation is done via JAVA and the performance is analyzed with various metrics.

Test Data Generation Using Optimization Algorithm: An Empirical Evaluation

This paper aims to design an approach for making an efficient fitness function for tests case generation based on distance from the goals. The designed function is given as an input to the genetic algorithm, and the result of the search process using the formed fitness function is evaluated in terms of time and the average number of test cases generated. This paper also investigates the effect of parameter setting such as the size of the initial population on the performance of genetic algorithm using the proposed fitness function. The experimental result shows that the proposed approach is both time and cost efficient in comparison with manual and random testing. It is also found that initial larger population size gives better results in comparison with low initial population.

Bio-inspired metaheuristics: evolving and prioritizing software test data

Software testing is both a time and resource-consuming activity in software development. The most difficult parts of software testing are the generation and prioritization of test data. Principally these two parts are performed manually. Hence introducing an automation approach will significantly reduce the total cost incurred in the software development lifecycle. A number of automatic test case generation (ATCG) and prioritization approaches have been explored. In this paper, we propose two approaches: (1) a pathspecific approach for ATCG using the following metaheuristic techniques: the genetic algorithm (GA), particle swarm optimization (PSO) and artificial bee colony optimization (ABC); and (2) a test case prioritization (TCP) approach using PSO. Based on our experimental findings, we conclude that ABC outperforms the GA and PSO-based approaches for ATC.G Moreover, the results for PSO on TCP arguments demonstrate biased applicability for both small and large test suites against random, reverse and unordered prioritization schemes. Therefore, we focus on conducting a comprehensive and exhaustive study of the application of metaheuristic algorithms in solving ATCG and TCP problems in software engineering.

Software effort estimation using machine learning techniques

Effort Estimation is a very important activity for planning and scheduling of software project life cycle in order to deliver the product on time and within budget. Machine learning techniques are proving very useful to accurately predict software effort values. This paper presents a review of various machine-learning techniques using in estimation of software project effort namely Artificial Neural Network, Fuzzy logic, Analogy estimation etc. Machine learning techniques consistently predicting accurate results because of its learning natures form previously completed projects. This paper summarizes that each technique has its own features and behave differently according to environment so no technique can be preferred over each other.

Aspect oriented software testing

Aspect oriented software engineering has become a promising domain to address the challenges faced by the procedural and object oriented software development. It has solved the problem of separation of concerns by modularizing the crosscutting concerns that helps in reducing the maintenance cost. AOSE has introduced new features which in turn produces new kind of errors. As the main goal of software development is to produce the best quality product and the quality of software is widely depends on how testing is performed. Hence, testing remains an inescapable task in aspect oriented software development. In this paper an outline of aspect-oriented software engineering, along with the research work done by various authors on testing methods used in the area of aspect-oriented paradigm is presented. It is found that unit testing, integration testing and mutation testing are widely used techniques in aspect oriented software testing. It is also predicted from the approaches that graph based coverage is widely used as compared with other approaches.

Computational intelligence based approaches to software reliability

Accurate software reliability prediction with a single universal software reliability growth model is very difficult. In this ρ aper we reviewed different models which uses computational intelligence for the prediction purpose and describe how these techniques outperform conventional statistical models. Parameters, efficacy measures with methodologies are concluded in tabular form.

Efficient parameter estimation of software reliability growth models using harmony search

The primary challenge of software reliability growth model is to find the unknown model parameters that are used to validate on software failure dataset. Though, numerical estimation technique plays a vital role in parameter estimation of software reliability growth models, they are not optimal as they suffer from constraints sucha as sample size, biasing, and initialisation of parameters. In this study, a parameter estimation of software reliability growth model that utilises a variant of harmony search is proposed. Extensive experiments are conducted on seven different software datasets of varying complexity. A robust experimental setup is developed employing an orthogonal array and Taguchi method. Two-fold performance comparisons are performed. First, the authors tested their proposed approach against Cuckoo search and numerical method (least square estimation) considering mean square error and Theils statistics as a quality measure. Second, the authors applied statistical tests are performed that demonstrate the superiority of their approach over the others. The underlying motivation to conduct this study is to motivate researchers to utilise their approach for a better estimation of model parameters.