Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userWhat Are We Really Testing in Mutation Testing for Machine Learning? A Critical Reflection
149
0
0.0
(
0
)
Added by
Annibale Panichella
Publication date
2021
fields
Informatics Engineering
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Mutation testing is a well-established technique for assessing a test suites quality by injecting artificial faults into production code. In recent years, mutation testing has been extended to machine learning (ML) systems, and deep learning (DL) in particular; researchers have proposed approaches, tools, and statistically sound heuristics to determine whether mutants in DL systems are killed or not. However, as we will argue in this work, questions can be raised to what extent currently used mutation testing techniques in DL are actually in line with the classical interpretation of mutation testing. We observe that ML model development resembles a test-driven development (TDD) process, in which a training algorithm (`programmer) generates a model (program) that fits the data points (test data) to labels (implicit assertions), up to a certain threshold. However, considering proposed mutation testing techniques for ML systems under this TDD metaphor, in current approaches, the distinction between production and test code is blurry, and the realism of mutation operators can be challenged. We also consider the fundamental hypotheses underlying classical mutation testing: the competent programmer hypothesis and coupling effect hypothesis. As we will illustrate, these hypotheses do not trivially translate to ML system development, and more conscious and explicit scoping and concept mapping will be needed to truly draw parallels. Based on our observations, we propose several action points for better alignment of mutation testing techniques for ML with paradigms and vocabularies of classical mutation testing.
rate research
Read More
Deep learning (DL) defines a new data-driven programming paradigm where the internal system logic is largely shaped by the training data. The standard way of evaluating DL models is to examine their performance on a test dataset. The quality of the test dataset is of great importance to gain confidence of the trained models. Using an inadequate test dataset, DL models that have achieved high test accuracy may still lack generality and robustness. In traditional software testing, mutation testing is a well-established technique for quality evaluation of test suites, which analyzes to what extent a test suite detects the injected faults. However, due to the fundamental difference between traditional software and deep learning-based software, traditional mutation testing techniques cannot be directly applied to DL systems. In this paper, we propose a mutation testing framework specialized for DL systems to measure the quality of test data. To do this, by sharing the same spirit of mutation testing in traditional software, we first define a set of source-level mutation operators to inject faults to the source of DL (i.e., training data and training programs). Then we design a set of model-level mutation operators that directly inject faults into DL models without a training process. Eventually, the quality of test data could be evaluated from the analysis on to what extent the injected faults could be detected. The usefulness of the proposed mutation testing techniques is demonstrated on two public datasets, namely MNIST and CIFAR-10, with three DL models.
In the field of mutation analysis, mutation is the systematic generation of mutated programs (i.e., mutants) from an original program. The concept of mutation has been widely applied to various testing problems, including test set selection, fault localization, and program repair. However, surprisingly little focus has been given to the theoretical foundation of mutation-based testing methods, making it difficult to understand, organize, and describe various mutation-based testing methods. This paper aims to consider a theoretical framework for understanding mutation-based testing methods. While there is a solid testing framework for general testing, this is incongruent with mutation-based testing methods, because it focuses on the correctness of a program for a test, while the essence of mutation-based testing concerns the differences between programs (including mutants) for a test. In this paper, we begin the construction of our framework by defining a novel testing factor, called a test differentiator, to transform the paradigm of testing from the notion of correctness to the notion of difference. We formally define behavioral differences of programs for a set of tests as a mathematical vector, called a d-vector. We explore the multi-dimensional space represented by d-vectors, and provide a graphical model for describing the space. Based on our framework and formalization, we interpret existing mutation-based fault localization methods and mutant set minimization as applications, and identify novel implications for future work.
Mutation testing is used to evaluate the effectiveness of test suites. In recent years, a promising variation called extreme mutation testing emerged that is computationally less expensive. It identifies methods where their functionality can be entirely removed, and the test suite would not notice it, despite having coverage. These methods are called pseudo-tested. In this paper, we compare the execution and analysis times for traditional and extreme mutation testing and discuss what they mean in practice. We look at how extreme mutation testing impacts current software development practices and discuss open challenges that need to be addressed to foster industry adoption. For that, we conducted an industrial case study consisting of running traditional and extreme mutation testing in a large software project from the semiconductor industry that is covered by a test suite of more than 11,000 unit tests. In addition to that, we did a qualitative analysis of 25 pseudo-tested methods and interviewed two experienced developers to see how they write unit tests and gathered opinions on how useful the findings of extreme mutation testing are. Our results include execution times, scores, numbers of executed tests and mutators, reasons why methods are pseudo-tested, and an interview summary. We conclude that the shorter execution and analysis times are well noticeable in practice and show that extreme mutation testing supplements writing unit tests in conjunction with code coverage tools. We propose that pseudo-tested code should be highlighted in code coverage reports and that extreme mutation testing should be performed when writing unit tests rather than in a decoupled session. Future research should investigate how to perform extreme mutation testing while writing unit tests such that the results are available fast enough but still meaningful.
Diversity has been proposed as a key criterion to improve testing effectiveness and efficiency.It can be used to optimise large test repositories but also to visualise test maintenance issues and raise practitioners awareness about waste in test artefacts and processes. Even though these diversity-based testing techniques aim to exercise diverse behavior in the system under test (SUT), the diversity has mainly been measured on and between artefacts (e.g., inputs, outputs or test scripts). Here, we introduce a family of measures to capture behavioural diversity (b-div) of test cases by comparing their executions and failure outcomes. Using failure information to capture the SUT behaviour has been shown to improve effectiveness of history-based test prioritisation approaches. However, history-based techniques require reliable test execution logs which are often not available or can be difficult to obtain due to flaky tests, scarcity of test executions, etc. To be generally applicable we instead propose to use mutation testing to measure behavioral diversity by running the set of test cases on various mutat
ReTest is a novel testing tool for Java applications with a graphical user interface (GUI), combining monkey testing and difference testing. Since this combination sidesteps the oracle problem, it enables the generation of GUI-based regression tests. ReTest makes use of evolutionary computing (EC), particularly a genetic algorithm (GA), to optimize these tests towards code coverage. While this is indeed a desirable goal in terms of software testing and potentially finds many bugs, it lacks one major ingredient: human behavior. Consequently, human testers often find the results less reasonable and difficult to interpret. This thesis proposes a new approach to improve the initial population of the GA with the aid of machine learning (ML), forming an ML-technique enhanced-EC (MLEC) algorithm. In order to do so, existing tests are exploited to extract information on how human testers use the given GUI. The obtained data is then utilized to train an artificial neural network (ANN), which ranks the available GUI actions respectively their underlying GUI components at runtime---reducing the gap between manually created and automatically generated regression tests. Although the approach is implemented on top of ReTest, it can be easily used to guide any form of monkey testing. The results show that with only little training data, the ANN is able to reach an accuracy of 82% and the resulting tests represent an improvement without reducing the overall code coverage and performance significantly.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
