Abstract:
Software testing is a critical phase in software development, ensuring the quality, reliability, and security of software systems. As software complexity increases, selecting an efficient testing approach becomes essential. This study compares manual and automated software testing to evaluate their effectiveness in identifying defects, efficiency in terms of time and cost, and overall accuracy in defect detection. While manual testing relies on human expertise for exploratory and usability testing, automated testing leverages specialized tools to improve efficiency and consistency. Despite their advantages, both methods have limitations, making their comparative study valuable for software engineers and testers. This research employs machine learning techniques such as Random Forest Classification, Logistic Regression Classification and Linear Regression to predict testing efficiency based on various factors, including experience and testing type. The study aims to provide a data-driven comparison of defect detection accuracy and efficiency, identifying scenarios where one approach outperforms the other. The findings will help software stakeholders make informed decisions about optimizing testing strategies, balancing accuracy, efficiency, and cost while overcoming challenges associated with both testing methodologies.
