Scholarlink

Boundary Value Analysis (BVA) is a black box testing method that uses input boundaries to create test cases. this technique, however, often results in a high number of test cases which can be which can be time-consuming and inefficient. Thus, there is rising demand for a smart automated method that limits the number of test cases to the most representative ones without compromising efficiency. a technique has been proposed Boundary Value Analysis with Multi-Layer Perceptron Adam optimizer (BVA-MLPAdam) is a method that has been created to optimize the software testing process by deeply reducing the test cases to a minimum while delivering comprehensive coverage of boundary-based input classes. The optimal test cases are automatically selected using an intelligent model of a neural network, excluding manual selection and thus reducing human error during the testing process. The fully automated process maximizes efficiency and delivers more accurate and reliable test results. this approach is particularly valuable in evaluating large and complex systems with lots of variables and boundary inputs when it is impractical to use human selection to pick test cases due to an excessively high amount of potential input combinations. Choosing test cases within the method adopted here is accomplished through a multi-layer permutation (MLP) neural network which has been tuned using the Adam algorithm to systematically pick the best representative and productive test cases. This technique covers the entire testing process from test case generation and optimization through execution to the improved, faster testing process with reduced testing time and higher quality and reliability of the software tested. The effectiveness of the hybrid technique was assessed using three metrics: Precision, Recall, and F1 Score. The technique had optimal results in all cases. The observed results for all selected metrics were 1.00 on next-date function and Iris Benchmark Dataset indicating excellent precision, coverage and efficacy in test case generation.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Developing a New Optimization-Based Black-Box Software Testing Tool Using a Neural Network Improved with Adam Optimizer and Boundary Value Analysis

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Nada N. Saleem,
  • Zainab M. Alshamaa

摘要

Boundary Value Analysis (BVA) is a black box testing method that uses input boundaries to create test cases. this technique, however, often results in a high number of test cases which can be which can be time-consuming and inefficient. Thus, there is rising demand for a smart automated method that limits the number of test cases to the most representative ones without compromising efficiency. a technique has been proposed Boundary Value Analysis with Multi-Layer Perceptron Adam optimizer (BVA-MLPAdam) is a method that has been created to optimize the software testing process by deeply reducing the test cases to a minimum while delivering comprehensive coverage of boundary-based input classes. The optimal test cases are automatically selected using an intelligent model of a neural network, excluding manual selection and thus reducing human error during the testing process. The fully automated process maximizes efficiency and delivers more accurate and reliable test results. this approach is particularly valuable in evaluating large and complex systems with lots of variables and boundary inputs when it is impractical to use human selection to pick test cases due to an excessively high amount of potential input combinations. Choosing test cases within the method adopted here is accomplished through a multi-layer permutation (MLP) neural network which has been tuned using the Adam algorithm to systematically pick the best representative and productive test cases. This technique covers the entire testing process from test case generation and optimization through execution to the improved, faster testing process with reduced testing time and higher quality and reliability of the software tested. The effectiveness of the hybrid technique was assessed using three metrics: Precision, Recall, and F1 Score. The technique had optimal results in all cases. The observed results for all selected metrics were 1.00 on next-date function and Iris Benchmark Dataset indicating excellent precision, coverage and efficacy in test case generation.