Ali Ghanbari
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Ali Ghanbari has a total of 7 co-authors. Here are the top co-authors by number of shared publications:
- Andrian Marcus
George Mason University, Fairfax, Virginia, United States of America
1 shared publication - Ben Greenman
University of Utah, Salt Lake City, USA
1 shared publication - Sasan Tavakkol
Department of Civil and Environmental Engineering University of Southern California Los Angeles California USA
1 shared publication - Shawn Burnham
Auburn University, Department of Computer Science and Software Engineering, Auburn, USA
1 shared publication - Shibbir Ahmed
Department of Computer Science, Texas State University, San Marcos, TX, USA
1 shared publication - Tommi Rose Fahy
Auburn University, Department of Computer Science and Software Engineering, Auburn, USA
1 shared publication
Latest Publications
- Open Access
- 3 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 132-137
ADEPT: A Unified Framework for Deep Learning Test Adequacy
- 3 October 2026
- Published by Association for Computing Machinery (ACM)
- p. 132-137
Open AccessADEPT: A Unified Framework for Deep Learning Test Adequacy
- Open Access
- 1 October 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Software Engineering
- Vol. 3 (ISSTA) , 3683-3704
Provably Lossless Acceleration of DNN Mutation Testing via Memoization
Mutation analysis has recently reemerged in the context of deep neural networks (DNNs) as a promising, but notoriously costly, approach for assessing test dataset adequacy. Existing techniques speed up DNN mutation testing through lossy approximations that trade efficiency for mutation score accuracy. This paper introduces Mure, the first provably lossless framework for accelerating DNN mutation testing via memoization. Mure is based on the idea that DNN mutants and the original model share substantial redundant computation, so during mutation testing, it executes only the mutated suffixes of each mutant and reuses the common prefix from the original model, which is computed only once. We give a formal account of memoized mutation testing, and prove that Mure is sound, i.e., it produces results equivalent to exhaustive vanilla mutation testing, and identify basic conditions under which speed-up is guaranteed. We have implemented Mure and evaluated it on 15 DNN models of various architectures, complexities, and sizes ranging from a few thousands to millions of parameters. This provides empirical evidence that Mure reduces the computational cost of mutation testing by 44.54%, on average. We also observed that while state-of-the-art techniques tend to yield higher acceleration (up to 88.97%, on average), they come at the cost of some error in mutation score. We further analyze the effect of mutation generation selection ratio on the effectiveness of Mure and observed predictable reductions in memoization opportunities with increasing the percentage of mutated neurons. We observed that Mure offers more than 20% speed-up even when as high as 5% of the neurons are mutated.
- 1 October 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Software Engineering
- Vol. 3 (ISSTA) , 3683-3704
Open AccessProvably Lossless Acceleration of DNN Mutation Testing via Memoization
Mutation analysis has recently reemerged in the context of deep neural networks (DNNs) as a promising, but notoriously costly, approach for assessing test dataset adequacy. Existing techniques speed up DNN mutation testing through lossy approximations that trade efficiency for mutation score accuracy. This paper introduces Mure, the first provably lossless framework for accelerating DNN mutation testing via memoization. Mure is based on the idea that DNN mutants and the original model share substantial redundant computation, so during mutation testing, it executes only the mutated suffixes of each mutant and reuses the common prefix from the original model, which is computed only once. We give a formal account of memoized mutation testing, and prove that Mure is sound, i.e., it produces results equivalent to exhaustive vanilla mutation testing, and identify basic conditions under which speed-up is guaranteed. We have implemented Mure and evaluated it on 15 DNN models of various architectures, complexities, and sizes ranging from a few thousands to millions of parameters. This provides empirical evidence that Mure reduces the computational cost of mutation testing by 44.54%, on average. We also observed that while state-of-the-art techniques tend to yield higher acceleration (up to 88.97%, on average), they come at the cost of some error in mutation score. We further analyze the effect of mutation generation selection ratio on the effectiveness of Mure and observed predictable reductions in memoization opportunities with increasing the percentage of mutated neurons. We observed that Mure offers more than 20% speed-up even when as high as 5% of the neurons are mutated.
- Open Access
- 11 September 2024
- Published by Association for Computing Machinery (ACM)
- p. 1669-1681
Decomposition of Deep Neural Networks into Modules via Mutation Analysis
Recently, several approaches have been proposed for decomposing deep neural network (DNN) classifiers into binary classifier modules to facilitate modular development and repair of such models. These approaches concern only the problem of decomposing classifier models, and some of them rely on the activation patterns of the neurons, thereby limiting their applicability. In this paper, we propose a DNN decomposition technique, named Incite, that uses neuron mutation to quantify the contributions of the neurons to a given output of a model. Then, for each model output, a subgraph induced by the nodes with highest contribution scores for that output are selected and extracted as a module. Incite is agnostic to the type of the model and the activation functions used in its construction, and is applicable to not just classifiers, but to regression models as well. Furthermore, the costs of mutation analysis in Incite has been reduced by heuristic clustering of neurons, enabling its application to models with millions of parameters. Lastly, Incite prunes away the neurons that do not contribute to the outcome of the modules, producing compressed, efficient modules. We have evaluated Incite using 16 DNN models for well-known classification and regression problems and report its effectiveness along combined accuracy (and MAE) of the modules, the overlap in model elements between the modules, and the compression ratio. We observed that, for classification models, Incite, on average, incurs 3.44% loss in accuracy, and the average overlap between the modules is 71.76%, while the average compression ratio is 1.89X. Meanwhile, for regression models, Incite, on average, incurs 18.56% gain in MAE, and the overlap between modules is 80.14%, while the average compression ratio is 1.83X.
- 11 September 2024
- Published by Association for Computing Machinery (ACM)
- p. 1669-1681
Open AccessDecomposition of Deep Neural Networks into Modules via Mutation Analysis
Recently, several approaches have been proposed for decomposing deep neural network (DNN) classifiers into binary classifier modules to facilitate modular development and repair of such models. These approaches concern only the problem of decomposing classifier models, and some of them rely on the activation patterns of the neurons, thereby limiting their applicability. In this paper, we propose a DNN decomposition technique, named Incite, that uses neuron mutation to quantify the contributions of the neurons to a given output of a model. Then, for each model output, a subgraph induced by the nodes with highest contribution scores for that output are selected and extracted as a module. Incite is agnostic to the type of the model and the activation functions used in its construction, and is applicable to not just classifiers, but to regression models as well. Furthermore, the costs of mutation analysis in Incite has been reduced by heuristic clustering of neurons, enabling its application to models with millions of parameters. Lastly, Incite prunes away the neurons that do not contribute to the outcome of the modules, producing compressed, efficient modules. We have evaluated Incite using 16 DNN models for well-known classification and regression problems and report its effectiveness along combined accuracy (and MAE) of the modules, the overlap in model elements between the modules, and the compression ratio. We observed that, for classification models, Incite, on average, incurs 3.44% loss in accuracy, and the average overlap between the modules is 71.76%, while the average compression ratio is 1.89X. Meanwhile, for regression models, Incite, on average, incurs 18.56% gain in MAE, and the overlap between modules is 80.14%, while the average compression ratio is 1.83X.
- Open Access
- 10 October 2022
- Published by Association for Computing Machinery (ACM)
- p. 1-4
Shibboleth: Hybrid Patch Correctness Assessment in Automated Program Repair
Test-based generate-and-validate automated program repair (APR) systems generate many patches that pass the test suite without fixing the bug. The generated patches must be manually inspected by the developers, a task that tends to be time-consuming, thereby diminishing the role of APR in reducing debugging costs. We present the design and implementation of a novel tool, named Shibboleth, for automatic assessment of the patches generated by test-based generate-and-validate APR systems. Shibboleth leverages lightweight static and dynamic heuristics from both test and production code to rank and classify the patches. Shibboleth is based on the idea that the buggy program is almost correct and the bugs are small mistakes that require small changes to fix and specifically the fix does not remove the code implementing correct functionality of the program. Thus, the tool measures the impact of patches on both production code (via syntactic and semantic similarity) and test code (via code coverage) to separate the patches that result in similar programs and that do not remove desired program elements. We have evaluated Shibboleth on 1,871 patches, generated by 29 Java-based APR systems for Defects4J programs. The technique outperforms state-of-the-art raking and classification techniques. Specifically, in our ranking data set, in 66% of the cases, Shibboleth ranks the correct patch in top-1 or top-2 positions and, in our classification data set, it achieves an accuracy and F1-score of 0.887 and 0.852, respectively, in classification mode. A demo video of the tool is available at https://bit.ly/3NvYJN8.
- 10 October 2022
- Published by Association for Computing Machinery (ACM)
- p. 1-4
Open AccessShibboleth: Hybrid Patch Correctness Assessment in Automated Program Repair
Test-based generate-and-validate automated program repair (APR) systems generate many patches that pass the test suite without fixing the bug. The generated patches must be manually inspected by the developers, a task that tends to be time-consuming, thereby diminishing the role of APR in reducing debugging costs. We present the design and implementation of a novel tool, named Shibboleth, for automatic assessment of the patches generated by test-based generate-and-validate APR systems. Shibboleth leverages lightweight static and dynamic heuristics from both test and production code to rank and classify the patches. Shibboleth is based on the idea that the buggy program is almost correct and the bugs are small mistakes that require small changes to fix and specifically the fix does not remove the code implementing correct functionality of the program. Thus, the tool measures the impact of patches on both production code (via syntactic and semantic similarity) and test code (via code coverage) to separate the patches that result in similar programs and that do not remove desired program elements. We have evaluated Shibboleth on 1,871 patches, generated by 29 Java-based APR systems for Defects4J programs. The technique outperforms state-of-the-art raking and classification techniques. Specifically, in our ranking data set, in 66% of the cases, Shibboleth ranks the correct patch in top-1 or top-2 positions and, in our classification data set, it achieves an accuracy and F1-score of 0.887 and 0.852, respectively, in classification mode. A demo video of the tool is available at https://bit.ly/3NvYJN8.