Software Development Effort Estimation based-on multiple classifier system and Lines of Code

Estimation of Software Development Effort from Requirements Based Complexity

The estimation of software development effort has been centralized mostly on the accuracy of estimates through dealing with heterogeneous datasets regardless of the fact that the software projects are inherently complex and uncertain. In particular, Analogy Based Estimation (ABE), as a widely accepted estimation method, suffers a great deal from the problem of inconsistent and non-normal datasets because it is a comparison-based method and the quality of comparisons strongly depends on the consistency of projects. In order to overcome this problem, prior studies have suggested the use of weighting methods, outlier elimination techniques and various types of soft computing methods. However the proposed methods have reduced the complexity and uncertainty of projects, the results are not still convincing and the methods are limited to a special domain of software projects, which causes the generalization of methods to be impossible. Localization of comparison and weighting processes through clustering of projects is the main idea behind this paper. A hybrid model is proposed in which the software projects are divided into several clusters based on key attributes (development type, organization type and development platform). A combination of ABE and Particle Swarm Optimization (PSO) algorithm is used to design a weighting system in which the project attributes of different clusters are given different weights. Instead of comparing a new project with all the historical projects, it is only compared with the projects located in the related clusters based on the common attributes. The proposed method was evaluated through three real datasets that include a total of 505 software projects. The performance of the proposed model was compared with other well-known estimation methods and the promising results showed that the proposed localization can considerably improve the accuracy of estimates. Besides the increase in accuracy, the results also certified that the proposed method is flexible enough to be used in a wide range of software projects.

Software cost and time estimation is the process of estimating the cost and time required to develop a software system. Software cost and time estimation supports the planning and tracking of software projects. Effectively controlling the expensive investment of software development is one of the important issues in software project management. Estimating software development cost with high precision is still a great challenge for project managers, because it allows for considerable financial and strategic planning. Software cost estimation refers to the predictions of the likely amount of effort, time, and staffing levels required to build a software system. A very helpful form of cost estimation is the one made at an early stage during a project, when the costing of the project is proposed for approval. However, estimates at the early stages of the development are the most difficult to obtain. In this paper a novel Constructive Cost Model (COCOMO) based on soft computing approach is proposed for software cost estimation. This model carries some of the desirable features of neural networks approach, such as learning ability and good interpretability, while maintaining the merits of the COCOMO model. Unlike the standard neural networks approach, the proposed model can be interpreted and validated by experts, and has good generalisation capability. The model deals effectively with imprecise and uncertain input and enhances the reliability of software cost estimates. From the experimental results, it was concluded that, by the proposed neural network model, the accuracy of cost estimation can be improved and the estimated cost can be very close to the actual cost.

Systematic literature review of machine learning based software development effort estimation models

An international association advancing the multidisciplinary study of informing systems. Founded in 1998, the Informing Science Institute (ISI) is a global community of academics shaping the future of informing science.

Prediction of software development is the key task for the effective management of any software industry. The accuracy and reliability of the prediction mechanisms used for the estimation of software development effort is also important. A series of experiments are conducted to gradually progress towards the improved accurate estimation of the software development effort. However, while conducting these experiments, it was found that the size of the training set was not sufficient to train a large and complex artificial neural network (ANN). To overcome the problem of the size of the available training data set, a novel multilayered architecture based on a neural network model is proposed. The accuracy of the proposed multi-layered model is assessed using different criteria, which proves the pre-eminence of the proposed model.

Existen dos fases principales en el uso de un modelo de estimacion: (1) se debe determinar si el modelo es adecuado para describir los datos observados (reales), eso es, la comprobacion de la adecuacion del modelo o verificacion del mismo; si este resultara adecuado, entonces (2) el modelo de estimacion se valida en su ambiente usando datos nuevos. Este articulo esta relacionado con la primera etapa. Se presenta una investigacion dirigida a la comparacion de Sistemas de Logica Difusa (SLD) personales. Estos SLD se derivan a partir de un experimento replicado con base en una muestra de diez desarrolladores, asi como en un proceso de desarrollo comun dentro de un entorno controlado. En seis de los diez casos, las pruebas de rango multiple de la Magnitud del Error Relativo (MER) por tecnica, muestran que la logica difusa es ligeramente mejor que la regresion simple. Estos resultados muestran que un SLD podria ser utilizado como alternativa para la estimacion del esfuerzo de desarrollo de software a nivel personal

Software development effort estimation (SDEE) is one of the main tasks in software project management. It is crucial for a project manager to efficiently predict the effort or cost of a software project in a bidding process, since overestimation will lead to bidding loss and underestimation will cause the company to lose money. Several SDEE models exist; machine learning models, especially neural network models, are among the most prominent in the field. In this study, four different neural network models: Multilayer Perceptron, General Regression Neural Network, Radial Basis Function Neural Network, and Cascade Correlation Neural Network are compared with each other based on: (1) predictive accuracy centered on the Mean Absolute Error criterion, (2) whether such a model tends to overestimate or underestimate, and (3) how each model classifies the importance of its inputs. Industrial datasets from the International Software Benchmarking Standards Group (ISBSG) are used to train and validate the four models. The main ISBSG dataset was filtered and then divided into five datasets based on the productivity value of each project. Results show that the four models tend to overestimate in 80percent of the datasets, and the significance of the model inputs varies based on the selected model. Furthermore, the Cascade Correlation Neural Network outperforms the other three models in the majority of the datasets constructed on the Mean Absolute Residual criterion.

The Software Cost Estimation (SCE) is one of most important issues in the cycle of development, management decision, and in the quality of software project. In the case of the lack of certainty of the exact cost for software projects' development, companies encounter with numerous challenges. Also, due to inaccurate cost estimates, making wrong decisions by project managers make irreparable damage. To reach this propos, the effective factors for developing software projects should be evaluated to ensure of the projects success. COCOMO model is the main model for SCE which acts based on criteria and quantities such as number of Line of Code (LOC) or the Function Point Analysis (FPA). Research in recent years has shown that COCOMO model has not a good performance in the SCE. In this paper, we studied SCE by using a hybrid of Genetic Algorithm (GA) and Artificial Bee Colony (ABC) which are Meta-Heuristic Algorithms. Test results show that proposed model, GA and ABC algorithms have less MRE errors values than the COCOMO model. Also, the hybrid model has better convergence comparing with the GA and ABC algorithms.

GA-based method for feature selection and parameters optimization for machine learning regression applied to software effort estimation

Background. Currently there are several definitions of measures that should represent the size of software functional requirements. These measures have gained a quite relevant role, since they are one of the few basis upon which effort estimation can be based. However, traditional Functional Size Measures do not take into account the amount and complexity of the elaboration required, concentrating instead on the amount of data accessed or moved. This is a problem, when it comes to effort estimation, since the amount and complexity of the required data elaborations affect the implementation effort, but are not adequately represented by the current measures (including the standardized ones).

Software cost estimation (SCE) is one of the critical activities in software project management. During the past decades various models have been proposed for SCE. However, developing accurate and useful models is limited in practice despite the considerable financial gain they could offer to software stakeholders. Traditional techniques, such as regression, by-analogy and machine learning, face the difficulty of handling the dynamic nature of the software process and the problematic nature of the public data available. This paper addresses the issue of SCE proposing an alternative approach that combines robust decision tree structures with fuzzy logic. Fuzzy decision trees are generated using the CHAID and CART algorithms in a systematic manner, while development effort is treated as the dependent variable against two subsets of factors: The first contains selected attributes from the ISBSG, COCOMO and DESHARNAIS datasets and the second contains a subset of the available factors that can be measured early in the development cycle. The association rules obtained from the trees are then merged and defuzzified through a Fuzzy Implication System (FIS). The fuzzy framework is utilized to perform effort estimations. Experimental results indicate that the proposed approach is promising as it yields quite accurate estimations in most dataset cases considered. Finally, our evaluation suggests that accurate estimations may be produced, even when using only a small set of factors that can be measured early in the development cycle, thus increasing the practical value of the proposed cost model.

Favorable planed management is the base of software project succeed. In this piece aim at the present situation of software project management, comprehensive use the thought of software engineering and project management, explore process and method of project schedule planning control. On the base of analysis software project schedule planning control difficulties, research based on PSP method. PSP follow adjustment, practice, measurement, comparison result and readjustment such a circulation and the continuous improvement of process, it is a process step by step for a long time. This paper studies the successful application of content; it can change insufficient plan in software project management process, the phenomenon of control adverse, improve comprehensive management ability and competitive power of project, creating Value for software company, improve high quality service for user.

The development of a successfull software depends upon the accurate estimation, as various factors are responsible for the overall assessment of any project. The Sofware Cost Estimation is the most complicated and challenging task in software industry. Many estimation models are introduced by the time, that concludes estimation is not a precise science and demanding of new methodologies should be proposed day by day. Software Cost Estimation of a project is vital to the acceptance or rejection of the development of software project. Various techniques have been introduced. This paper enlightens two categories of Software Cost Estimation techniques ie Algorithimic and Non Algorithimic. This paper evaluates the Algorithmic Models such as SLIM, COCOMO, Function Points and Non Algorithimic Models such as Analogy, Expert Judgement which are used to estimate software costs. Poor planning frequently leads to project failure and considerable outcomes for the project team. Software project managers should be aware of the increasing of project failures. In this paper, several existing methods for software cost estimation are illustrated and their aspects will be discussed.

Software Requirements Changes is a typical phenomenon in any software development project. Restricting incoming changes might cause user dissatisfaction and allowing too many changes might cause delay in project delivery. Moreover, the acceptance or rejection of the change requests become challenging for software project managers when these changes are occurred in Software Development Phase. Where in Software Development Phase software artifacts are not in consistent state such as: some of the class artifacts are Fully Developed, some are Half Developed, some are Major Developed, some are Minor Developed and some are Not Developed yet. However, software effort estimation and change impact analysis are the two most common techniques which might help software project managers in accepting or rejecting change requests during Software Development Phase. The aim of this research is to develop a new software change effort estimation model which helps software project manager in estimating the effort for software Requirement Changes during Software Development Phase. Thus, this research has analyzed the existing effort estimation models and change impact analysis techniques for Softwrae Development Phase from the literature and proposed a new software change effort estimation model by combining change impact analysis technique with effort estimation model. Later, the new proposed model has been evaluated by selecting four small size software projects as case selections in applying experimental approch. The experiment results show that the overall Mean Magnitude Relative Error value produced by the new proposed model is under 25%. Hence it is concluded that the new proposed model is applicable in estimating the amount of effort for requirement changes during SDP.