Software belief reliability growth model incorporating change point and imperfect debugging based on uncertain...

Area of Science:

• Software Engineering
• Reliability Engineering
• Uncertainty Quantification

Background:

• Software reliability growth models (SRGMs) estimate reliability, faults, and release times.
• Imperfect debugging and changing testing strategies introduce epistemic uncertainty.
• Existing models struggle to effectively manage this uncertainty.

Purpose of the Study:

• To propose a novel software belief reliability growth model (SBRGM).
• To effectively handle epistemic uncertainty in software reliability.
• To improve accuracy by incorporating imperfect debugging and change points.

Main Methods:

• Utilized uncertain differential equations for modeling.
• Incorporated imperfect debugging and change point analysis via belief reliability theory.
• Employed the least squares method for parameter estimation and derivative principles for change point calculation.
• Validated the model using three real-world datasets and Python 3.10.

Main Results:

• The proposed SBRGM demonstrates increased accuracy compared to previous models.
• The model effectively manages epistemic uncertainty inherent in software development.
• Parameter estimation and change point calculation were successfully performed.

Conclusions:

• The SBRGM offers a more flexible and realistic approach to software reliability assessment.
• This research provides a valuable tool for understanding and mitigating uncertainty in software projects.
• The findings contribute to more reliable software development practices.