Design rework includes unnecessary repetition in design tasks to correct design problems. Resolving design matters in advance, through in-depth understanding of the design planning and rework issues and development of effective predictive tools could contribute to higher business profit margins and a faster product time-to-market. This research aims to develop three novel and structured methods to predict the design rework occurrence and effort at the very early design stage, which may otherwise remain undiscovered until the testing and refinement phase. The major contribution obtained from the Design Rework Probability of Occurrence Estimation method, DRePOE, is the development of design rework drivers. The developed drivers have been synthesised with data from interview results, direct observations, and archival records obtained from eleven world-class aerospace and automotive components manufacturers. To predict the probability of occurrence, the individual score of each driver was compared against historical records utilising the analogy-based method. The Design Rework Effort Estimation method, DREE, was developed to interconnect functional structures and identify failure relationships among components. A significant contribution of The DREE method is its capability to assess the design rework effort at the component level under the worst-case scenario. Next a Prioritisation Design by Design Rework Effort Based method, PriDDREB, was developed to provide a tool to forecast the maximum design rework given the constraint. This method provides a tool to determine and prioritise the components that may require a significant design rework effort. The three methods developed were validated with an automotive water pump, a turbocharger, and a McPherson strut suspension system in accordance with the validation square method. It is demonstrated that DRePOE, DREE, PriDDREB methods can offer the product design team a means to predict the probability of design rework occurrence and assess the required effort during the testing and refinement phase at the very early design phase.

This paper analyses and develops models for programmer participation and effort estimation in open source software projects. This has not yet been a centre of research, although any results would be of high importance for assessing the efficiency of this model and for various decision-makers. In this paper, a case study is used for hypotheses generation regarding manpower function and effort modeling, then a large data set retrieved from a project repository is used to test these hypotheses. The main results are that Norden-Rayleigh-based approaches need to be complemented to account for the addition of new features during the lifecycle to be usable in this context, and that programmer-participation based effort models show significantly less effort than those based on output metrics like lines-of-code. (author's abstract) / Series: Working Papers on Information Systems, Information Business and Operations

This master thesis is made in 2001 at Blekinge Institute of Technology and Symbian, which is a software company in Ronneby, Sweden. The purpose of the thesis is to find a suitable prediction and estimation model for the test effort. To do this, we have studied the State of the Art in cost/effort estimation and fault prediction. The conclusion of this thesis is that it is hard to make a general proposal, which is applicable for all organisations. For Symbian we have proposed a model based on use and test cases to predict the test effort.

It is important to manage iterative projects in a way to maximize quality and minimize cost. To achieve high quality, accurate project estimates are of high importance. It is challenging to predict the effort that is required to perform test activities in an iterative development. If testers put extra effort in testing then schedule might be delayed, however, if testers spend less effort then quality could be affected. Currently there is no model for test effort prediction in iterative development to overcome such challenges. This paper introduces and validates a dynamic Bayesian network to predict test effort in iterative software development. In this research work, the proposed framework is evaluated in a number of ways: First, the framework behavior is observed by considering different parameters and performing initial validation. Then secondly, the framework is validated by incorporating data from two industrial projects. The accuracy of the results has been verified through different prediction accuracy measurements and statistical tests. The results from the verification confirmed that the framework has the ability to predict test effort in iterative projects accurately.

Background: Software Development Effort Estimation is a process that focuses on estimating the required effort to develop a software project with a minimal budget. Estimating effort includes interpretation of required manpower, resources, time and schedule. Project managers are responsible for estimating the required effort. A model that can predict software development effort efficiently comes in hand and acts as a decision support system for the project managers to enhance the precision in estimating effort. Therefore, the context of this study is to increase the efficiency in estimating software development effort. Objective: The main objective of this thesis is to identify an effective ensemble method to build and implement it, in estimating software development effort. Apart from this, parameter tuning is also implemented to improve the performance of the model. Finally, we compare the results of the developed model with the existing models. Method: In this thesis, we have adopted two research methods. Initially, a Literature Review was conducted to gain knowledge on the existing studies, machine learning techniques, datasets, ensemble methods that were previously used in estimating Software Development Effort. Then a controlled Experiment was conducted in order to build an ensemble model and to evaluate the performance of the ensemble model for determining if the developed model has a better performance when compared to the existing models.   Results: After conducting literature review and collecting evidence, we have decided to build and implement stacked generalization ensemble method in this thesis, with the help of individual machine learning techniques like Support vector regressor (SVR), K-Nearest Neighbors regressor (KNN), Decision Tree Regressor (DTR), Linear Regressor (LR), Multi-Layer Perceptron Regressor (MLP) Random Forest Regressor (RFR), Gradient Boosting Regressor (GBR), AdaBoost Regressor (ABR), XGBoost Regressor (XGB). Likewise, we have decided to implement Randomized Parameter Optimization and SelectKbest function to implement feature section. Datasets like COCOMO81, MAXWELL, ALBERCHT, DESHARNAIS were used. Results of the experiment show that the developed ensemble model performs at its best, for three out of four datasets. Conclusion: After evaluating and analyzing the results obtained, we can conclude that the developed model works well with the datasets that have continuous, numeric type of values. We can also conclude that the developed ensemble model outperforms other existing models when implemented with COCOMO81, MAXWELL, ALBERCHT datasets.

<p>Saab Ericsson Space AB develops products for space for a predetermined price. Since the price is fixed, it is crucial to have a reliable prediction model to estimate the effort needed to develop the product. In general software effort estimation is difficult, and at the software department this is a problem.</p><p>By analyzing metrics, collected from former projects, different prediction models are developed to estimate the number of person hours a software project will require. Models for predicting the effort before a project begins is first developed. Only a few variables are known at this state of a project. The models developed are compared to a current model used at the company. Linear regression models improve the estimate error with nine percent units and nonlinear regression models improve the result even more. The model used today is also calibrated to improve its predictions. A principal component regression model is developed as well. Also a model to improve the estimate during an ongoing project is developed. This is a new approach, and comparison with the first estimate is the only evaluation.</p><p>The result is an improved prediction model. There are several models that perform better than the one used today. In the discussion, positive and negative aspects of the models are debated, leading to the choice of a model, recommended for future use.</p>

Software project managers usually use benchmarking repositories to estimate effort, cost, and duration of the software development which will be used to appropriately plan, monitor and control the project activities. In addition, precision of benchmarking repositories is a critical factor in software effort estimation process which plays subsequently a critical role in the success of the software development project. In order to construct such a precise benchmarking data repository, it is important to have defined benchmarking data attributes and data characteristics and to have collected project data accordingly. On the other hand, studies show that data characteristics of benchmark data sets have impact on generalizing the studies which are based on using these datasets. Quality of data repository is not only depended on quality of collected data, but also it is related to how these data are collected. In this thesis, a benchmarking methodology is proposed for organizations to collect benchmarking data for effort estimation purposes. This methodology consists of three main components: benchmarking measures, benchmarking data collection processes, and benchmarking data collection tool. In this approach results of previous studies from the literature were used too. In order to verify and validate the methodology project data were collected in two middle size software organizations and one small size organization by using automated benchmarking data collection tool. Also, effort estimation models were constructed and evaluated for these projects data and impact of different characteristics of the projects was inspected in effort estimation models.

Saab Ericsson Space AB develops products for space for a predetermined price. Since the price is fixed, it is crucial to have a reliable prediction model to estimate the effort needed to develop the product. In general software effort estimation is difficult, and at the software department this is a problem. By analyzing metrics, collected from former projects, different prediction models are developed to estimate the number of person hours a software project will require. Models for predicting the effort before a project begins is first developed. Only a few variables are known at this state of a project. The models developed are compared to a current model used at the company. Linear regression models improve the estimate error with nine percent units and nonlinear regression models improve the result even more. The model used today is also calibrated to improve its predictions. A principal component regression model is developed as well. Also a model to improve the estimate during an ongoing project is developed. This is a new approach, and comparison with the first estimate is the only evaluation. The result is an improved prediction model. There are several models that perform better than the one used today. In the discussion, positive and negative aspects of the models are debated, leading to the choice of a model, recommended for future use.

Context Web engineering is a new research field which utilizes engineering principles to produce quality web applications. Web applications have become more complex with the passage of time and it's quite difficult to analyze the web metrics for the estimation due to a wide range of web applications. Correct estimates for web development effort play a very important role in the success of large-scale web development projects. Objectives In this study I investigated size metrics and cost drivers used by web companies for early web cost estimation. I also aim to get validation through industrial interviews and web quote form. This form is designed based on most frequently occurring metrics after analyzing different companies. Secondly, this research aims to revisit previous work done by Mendes (a senior researcher and contributor in this research area) to validate whether early web cost estimation trends are same or changed? The ultimate goal is to help companies in web cost estimation. Methods First research question is answered by conducting an online survey through 212 web companies and finding their web predictor forms (quote forms). All companies included in the survey used Web forms to give quotes on Web development projects based on gathered size and cost measures. The second research question is answered by finding most occurring size metrics from the results of Survey 1. List of size metrics are validated by two methods: (i) Industrial interviews are conducted with 15 web companies to validate results of the first survey (ii) a quote form is designed using validated results from industrial interviews and quote form sent to web companies around the world to seek data on real Web projects. Data gathered from Web projects are analyzed using CBR tool and results are validated with Industrial interview results along with Survey 1.  Final results are compared with old research to justify answer of third research question whether size metrics have been changed. All research findings are contributed to Tukutuku research benchmark project. Results “Number of pages/features” and “responsive implementation” are top web size metrics for early Web cost estimation. Conclusions. This research investigated metrics which can be used for early Web cost estimation at the early stage of Web application development. This is the stage where the application is not built yet but just requirements are being collected and an expected cost estimation is being evaluated. List of new metrics variable is concluded which can be added in Tukutuku project.

Made available in DSpace on 2014-06-12T15:49:48Z (GMT). No. of bitstreams: 1 license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2009 / Em mercados competitivos como, por exemplo, o de celulares, empresas de software que liberam produtos com baixa qualidade podem rapidamente perder os seus clientes. A fim de evitar esse problema, essas empresas devem garantir que a qualidade dos produtos atenda a expectativa de seus clientes. Nesse contexto, testes é uma das atividades mais utilizadas para se tentar melhorar a qualidade de um software. Além disso, o resultado da atividade de teste está sendo considerado tão importante que em muitos casos é preferível alocar equipes exclusivamente para exercer atividades de teste. Essas equipes de teste devem ser capazes de estimar o esforço exigido para exercer as suas atividades dentro do prazo ou para solicitar mais recursos ou negociar prazos quando necessário. Na prática, as consequências de se ter estimativas ruins são onerosas para a organização: redução de escopo, atraso nas entregas ou horas extras de trabalho. O impacto dessas consequências é ainda maior em se tratando de execução manual de testes. Visando uma melhor forma de estimar esforço de execução manual de casos de teste funcionais, esta pesquisa propõe e valida uma medida de tamanho de teste e de complexidade de execução baseada nas próprias especificações dos testes, bem como um método de medição para a métrica proposta. Além disso, diversos estudos de caso, survey e experimentos foram realizados para avaliar o impacto desse trabalho. Durante esses estudos, verificamos uma melhoria significativa proporcionada por nossa abordagem na precisão das estimativas de esforço de execução de testes manuais. Também identificamos fatores de custo relacionados a atividades de execução manual de testes utilizando julgamento de especialistas. O efeito desses fatores foram investigados através da execução de experimentos controlados, onde pudemos constatar que apenas alguns dos fatores identificados tiveram efeito significativo. Por fim, diversas ferramentas de suporte foram desenvolvidas durante essa pesquisa, incluindo a automação das estimativas de esforço de execução de testes a partir de especificações de testes escritas em linguagem natural