• The KIPS Transactions:PartD
• /
• v.8D no.5
• /
• pp.581-586
• /
• 2001

An increasingly important facet of software development is the ability to estimated the associated coast and effort of development early in the development life cycle. In spite of the most generally sued procedures for estimation of the software development effort and cost were linear regression analysis. As a result of the software complexity and various development environments, the software effort and cost estimates that are grossly inaccurate. The application of nonlinear methods hold the greatest promise for achieving this objects. Therefore this paper presents an RBF (radial basis function) network model that is able to represent the nonlinear relation for software development effort, The research describes appropriate RBF network modeling in the context of a case study for 24 software development projects. Also, this paper compared the RBF network model with a regression analysis model. The RBF network model is the most accuracy of all.

• The KIPS Transactions:PartD
• /
• v.16D no.5
• /
• pp.729-736
• /
• 2009

The accurate estimation of software development cost is important to a successful development in software engineering. Until recent days, the model using regression analysis based on statistical algorithm and machine learning method have been used. However, this paper estimates the software cost using support vector regression, a sort of machine learning technique. Also, it finds the best set of optimized parameters applying genetic algorithm. The proposed GA-SVR model outperform some recent results reported in the literature.

• Korean Management Science Review
• /
• v.19 no.2
• /
• pp.191-204
• /
• 2002

As the portion of information systems (IS) budget to the total government budget becomes greater, the cost estimation of IS development and maintenance projects is recognized as one of the most important problems to be resolved for scientific and efficient management of IS budget. Since IS budget makes much effect on the delivery time, quality and productivity of IS projects, the exact cost estimation is also necessary for the successful accomplishment of IS projects. The primary concern in the cost estimation of IS projects is software cost estimation, which requires the measurement of the size of softwares. There are two methods for sizing software : line-of-code approach, function point model. In this paper, we propose a function-point-based model for estimating software cost. The proposed model is derived by collecting about fifty domestic IT projects in public sector and analyzing their relationship between cost drivers and development effort. Since the proposed model is developed by simplifying the function point model that can be used only when detailed user requirements are specified, it can be also applied at project planning and budgeting phase.

• Journal of Korea Society of Industrial Information Systems
• /
• v.5 no.3
• /
• pp.18-29
• /
• 2000

This study presents a cost of efforts estimation model under the environment of developing a software using component software package tools. The approach taken was to drive from variety of sources in an attempt to identify the most significant factors. These sources ranged from already existing cost models like COTS integration cost and COCOMO models to information gathered in a data collection survey. Once the candidate drivers had been identified, the next step was to interview with the experts who had been experienced more than 5 years in component development area to identify the most significant driving factors. From those selected drivers, I established the Cost Estimation Model which is suitable for the developing a software using component software package tools by applying the general from of the well-know COCOMO software cost estimation model. To established the best fit in Korean Software industry, I used Regression statistical analysis with 31 data collections.

• Journal of the Korea Society of Computer and Information
• /
• v.14 no.12
• /
• pp.225-232
• /
• 2009

Function Point Method have been applied to measure software size estimation in industry because it supports to estimate the software's size by user's view not developer's. However, the current function point method has some problems for example complexity's upper limit etc. So, In this paper, an enhanced function point model. Micro FP model, was suggested. Using this model, software effort estimation can be more efficiently because this model has some regression equation. This model specially can be applied to estimate in detail the large application system's size Analysis results show that measured software size by this Micro FP model has the advantage with more correlative between the one of LOC, as of 10 applications operated in an large organization.

• Journal of Information Processing Systems
• /
• v.8 no.4
• /
• pp.621-652
• /
• 2012

Software Estimations provide an inclusive set of directives for software project developers, project managers, and the management in order to produce more realistic estimates based on deficient, uncertain, and noisy data. A range of estimation models are being explored in the industry, as well as in academia, for research purposes but choosing the best model is quite intricate. Estimation by Analogy (EbA) is a form of case based reasoning, which uses fuzzy logic, grey system theory or machine-learning techniques, etc. for optimization. This research compares the estimation accuracy of some conventional data mining models with a hybrid model. Different data mining models are under consideration, including linear regression models like the ordinary least square and ridge regression, and nonlinear models like neural networks, support vector machines, and multivariate adaptive regression splines, etc. A precise and comprehensible predictive model based on the integration of GRA and regression has been introduced and compared. Empirical results have shown that regression when used with GRA gives outstanding results; indicating that the methodology has great potential and can be used as a candidate approach for software effort estimation.

• Journal of the Korea Society of Computer and Information
• /
• v.4 no.2
• /
• pp.61-69
• /
• 1999

In this study, the productivity factors of software maintenance are categorized into the three areas such as maintainer characteristics, software technical characteristics and maintenance environment characteristics. And the various regression analyses are performed for searching the most significant model by using actual project's data. At conclusion, an linear model including new variables derived from factor analysis to estimate the effort for software maintenance project is suggested.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.13 no.4
• /
• pp.99-106
• /
• 2013

Generally, software development schedule compression is decided within 75% of nominal schedule. However, there are some difficulties when we judge the possibility of development schedule compression because a nominal schedule has not been definitely. Therefore, this paper investigates various nominal schedule estimation models and decides the optimum range. Basing on the performance of ISBSG Release 8,614 projects are extracted from the actual 1,595 projects. We presented development effort estimation model from those data. Also, we derived the development schedule model from the development effort. When you apply the proposed model, you will be able to estimate development effort and schedule required for the development more actually.

• The KIPS Transactions:PartD
• /
• v.11D no.7 s.96
• /
• pp.1483-1490
• /
• 2004

To accomplish a project successfuly, we have to estimate develpment effort accurately. But, development effort is different to software size and operation environment. Usually, we made use of function point for estimating development effort. In this paper. we make use of 789 project data. It is related to development projects in 1990`s. We investigate the variable affecting development effort. Also, we exedcute multiple liner regression analysis for looking linear relation about variables. We find the regression equation for multistage by dividing PDR that influ-enced development effort step by step.

• Journal of the Korea Computer Industry Society
• /
• v.7 no.3
• /
• pp.191-198
• /
• 2006

Area of software measurement is active more than thirty years. There is a huge collection of researches but still no concrete software development effort, duration and cost estimation model. The data sets used to conduct previous studies in the duration estimation model are often small and not too recent, these types of models should not be apply in recent projects that have complex architecture and various development environment. Therefore, Oligny et al. presents empirical models that predict software project duration in accordance with project platform based on project effort using the log data transformation. These models are based on the analysis of 396 project data provided by release 4 of the ISBSG Benchmark. Applying Oligny et al.'s models to 534 project data provided release 6 of the ISBSG Benchmark, the project duration is affected by development type more than development platform. Therefore, This paper presents the model of duration estimation according to development type. This paper proves the duration is more affected by development type than development platform. And, The model according to development type is more adequate for duration estimation.