Improved Software Reliability Prediction by Using Model Stacking and Averaging

Karaomer, Rabia Burcu
Yet, Barbaros
Software reliability is an important factor for the success of a software project. Accurate modelling of software reliability enables estimation of remaining defects, the timing of deployment and required future effort. These factors contribute to successful planning of project schedule and resources. A number of software reliability prediction models have been proposed, each with different assumptions regarding software defect introduction and discovery. The performances of these models differ depending on the properties of the software project they are applied. Model averaging and stacking techniques offer flexible approaches for combining the predictions of different models based on observed data. In this study, we use model stacking and averaging approaches to combine the predictions of four well-known Non-Homogeneous Poisson Process (NHPP) software reliability models. These models have different assumptions with respect to failure rate, residual defects and the overall reliability of the software being investigated. We evaluate these techniques in simulated experiments and then apply the techniques to defect data collected from four software projects with different characteristics. Our results show that stacking and averaging approaches provide a robust approach with consistently high-performance results for both simulated defect experiments and actual defect data, whereas the performance of individual prediction models varies between different projects.


A unification model and tool support for software functional size measurement methods
Efe, Pınar; Demirörs, Onur; Department of Information Systems (2006)
Software size estimation/measurement has been the objective of a lot of research in the software engineering community due to the need of reliable size estimates. FSM Methods have become widely used in software project management to measure the functional size of software since its first publication, late 1970s. Although all FSM methods measure the functional size by quantifying the FURs, each method defined its own measurement process and metric. Therefore, a piece of software has several functional sizes ...
An Effort Prediction Model Based on BPM Measures for Process Automation
Aysolmaz, Banu; Iren, Deniz; Demirörs, Onur (2013-06-18)
BPM software automation projects require different approaches for effort estimation for they are developed based on business process models rather than traditional requirements analysis outputs. In this empirical research we examine the effect of various measures for BPMN compliant business process models on the effort spent to automate those models. Although different measures are suggested in the literature, only a few studies exist that relate these measures to effort estimation. We propose that differen...
Investigation of the effects of reuse on software quality inan industrial setting
Deniz, Berkhan; Bilgen, Semih; Department of Electrical and Electronics Engineering (2013)
Software reuse is a powerful tool in order to reduce development and maintenance time and cost. Any software life cycle product can be reused, not only fragments of source code. A high degree of reuse correlates with a low defect density. In the literature, many theoretical and empirical researches have examined the relationship of software reuse and quality. In this thesis, the effects of reuse on software quality are investigated in an industrial setting. Throughout this study, we worked with Turkey’s lea...
Exploration of an Error Prevention Model for COSMIC Functional Size Measurement Method
Salmanoğlu, Murat; Demirörs, Onur (2012-10-19)
The reliability of Functional size is crucial for effort and cost estimation in software development projects. However, estimation results may differ because of the errors made in the measurement process. This research proposes an error prevention model for COSMIC Functional Size Measurement method to increase the reliability of the measurements. The model has two main suggestions; first, to use the types of the functional processes during the measurement to prevent errors, and the second, to use a cardinal...
Use of project similarity for software development time estimation
Taştekin, Semra Yilmaz; Bilgen, Semih; Erten, Murat; Department of Information Systems (2013)
Accurate development time estimation is crucial for project management in general, and critical for software intensive systems projects, in particular. Before beginning the project, little information is available for development details. Therefore, development time may not be estimated correctly. If data on previous projects in the same domain is available, this can be used for development time estimations. At the beginning of the project, requirements are defined and requirements specification document is...
Citation Formats
R. B. Karaomer, B. Yet, and O. CHOUSEİNOGLOU, “Improved Software Reliability Prediction by Using Model Stacking and Averaging,” 2019, Accessed: 00, 2020. [Online]. Available: