Regression testing is a testing activity that is performed to provide confidence that changes do not harm the existing behaviour of the software. Test suites tend to grow in size as software evolves, often making it too costly to execute entire test suites. A number of different approaches have been studied to maximize the value of the accrued test suite: minimization, selection and prioritization. Test suite minimization seeks to eliminate redundant test cases in order to reduce the number of tests to run. Test case selection seeks to identify the test cases that are relevant to some set of recent changes. Test case prioritization seeks to order test cases in such a way that early fault detection is maximized. This paper surveys each area of minimization, selection and prioritization technique and discusses open problems and potential directions for future research.This survey aims to collect and consider papers that deal with three regression testing techniques: test suite minimization, RTS and test case prioritization. Our intention is not to undertake a systematic review, but rather to provide a broad state-of-the-art view on these related fields. Many different approaches have been proposed to aid regression testing, which has resulted in a body of literature that is spread over a wide variety of domains and publication venues. The majority of surveyed literature has been published in the software engineering domain, and especially in the software testing and software maintenance literature. However, the regression testing literature also overlaps with those of programming language analysis, empirical software engineering and software metrics.Therefore, the paper selection criteria on which this survey is based are the problems considered in papers, while focusing on the specific topics of minimization, selection and prioritization. The formal definitions of these problems are presented in Section 2.2. The selected papers are listed in the Appendix. Fast abstracts and short papers have been excluded. BACKGROUNDThis section introduces the basic concepts and definitions that form a nomenclature of regression testing and minimization, selection and prioritization techniques. Definition 1 (Test Suite Minimization Problem)Given: A test suite, T , a set of test requirements {r 1 , . . .,r n }, that must be satisfied to provide the desired 'adequate' testing of the program, and subsets of T , T 1 , . . ., T n , one associated with each of the r i s such that any one of the test cases t j belonging to T i can be used to achieve requirement r i .Problem: Find a representative set, T , of test cases from T that satisfies all r i s.The testing criterion is satisfied when every test requirement in {r 1 , . . .,r n } is satisfied. A test requirement, r i , is satisfied by any test case, t j , that belongs to the T i , a subset of T . Therefore, the representative set of test cases is the hitting set of the T i s. Furthermore, in order to maximize the effect of minimization, T should be the minimal hitting set of the T i ...

Deep Learning (DL) systems are rapidly being adopted in safety and security critical domains, urgently calling for ways to test their correctness and robustness. Testing of DL systems has traditionally relied on manual collection and labelling of data. Recently, a number of coverage criteria based on neuron activation values have been proposed. These criteria essentially count the number of neurons whose activation during the execution of a DL system satisfied certain properties, such as being above predefined thresholds. However, existing coverage criteria are not sufficiently fine grained to capture subtle behaviours exhibited by DL systems. Moreover, evaluations have focused on showing correlation between adversarial examples and proposed criteria rather than evaluating and guiding their use for actual testing of DL systems. We propose a novel test adequacy criterion for testing of DL systems, called Surprise Adequacy for Deep Learning Systems (SADL), which is based on the behaviour of DL systems with respect to their training data. We measure the surprise of an input as the difference in DL system's behaviour between the input and the training data (i.e., what was learnt during training), and subsequently develop this as an adequacy criterion: a good test input should be sufficiently but not overtly surprising compared to training data. Empirical evaluation using a range of DL systems from simple image classifiers to autonomous driving car platforms shows that systematic sampling of inputs based on their surprise can improve classification accuracy of DL systems against adversarial examples by up to 77.5% via retraining.

SUMMARYRegression testing is a testing activity that is performed to provide confidence that changes do not harm the existing behaviour of the software. Test suites tend to grow in size as software evolves, often making it too costly to execute entire test suites. A number of different approaches have been studied to maximize the value of the accrued test suite: minimization, selection and prioritization. Test suite minimization seeks to eliminate redundant test cases in order to reduce the number of tests to run. Test case selection seeks to identify the test cases that are relevant to some set of recent changes. Test case prioritization seeks to order test cases in such a way that early fault detection is maximized. This paper surveys each area of minimization, selection and prioritization technique and discusses open problems and potential directions for future research. Copyright © 2010 John Wiley & Sons, Ltd.

Testing involves examining the behaviour of a system in order to discover potential faults. Given an input for a system, the challenge of distinguishing the corresponding desired, correct behaviour from potentially incorrect behavior is called the "test oracle problem". Test oracle automation is important to remove a current bottleneck that inhibits greater overall test automation. Without test oracle automation, the human has to determine whether observed behaviour is correct. The literature on test oracles has introduced techniques for oracle automation, including modelling, specifications, contract-driven development and metamorphic testing. When none of these is completely adequate, the final source of test oracle information remains the human, who may be aware of informal specifications, expectations, norms and domain specific information that provide informal oracle guidance. All forms of test oracles, even the humble human, involve challenges of reducing cost and increasing benefit. This paper provides a comprehensive survey of current approaches to the test oracle problem and an analysis of trends in this important area of software testing research and practice.

Abstract. The aim of Search Based Software Engineering (SBSE) research is to move software engineering problems from human-based search to machine-based search, using a variety of techniques from the metaheuristic search, operations research and evolutionary computation paradigms. The idea is to exploit humans' creativity and machines' tenacity and reliability, rather than requiring humans to perform the more tedious, error prone and thereby costly aspects of the engineering process. SBSE can also provide insights and decision support. This tutorial will present the reader with a step-by-step guide to the application of SBSE techniques to Software Engineering. It assumes neither previous knowledge nor experience with Search Based Optimisation. The intention is that the tutorial will cover sufficient material to allow the reader to become productive in successfully applying search based optimisation to a chosen Software Engineering problem of interest. IntroductionSearch Based Software Engineering (SBSE) is the name given to a body of work in which Search Based Optimisation is applied to Software Engineering. This approach to Software Engineering has proved to be very successful and generic. It has been a subfield of software engineering for ten years [45], the past five of which have been characterised by an explosion of interest and activity [48]. New application areas within Software Engineering continue to emerge and a body of empirical evidence has now accrued that demonstrates that the search based approach is definitely here to stay. SBSE seeks to reformulate Software Engineering problems as 'search problems' [45,48]. This is not to be confused with textual or hypertextual searching. Rather, for Search Based Software Engineering, a search problem is one in which optimal or near optimal solutions are sought in a search space of candidate solutions, guided by a fitness function that distinguishes between better and worse solutions. The term SBSE was coined by Harman and Jones [45] in 2001, which was the first paper to advocate Search Based Optimisation as a general approach to Software Engineering, though there were other authors who had previously applied search based optimisation to aspects of Software Engineering.SBSE has been applied to many fields within the general area of Software Engineering, some of which are already sufficiently mature to warrant their own surveys. For example, there are surveys and overviews, covering SBSE for requirements [111], design [78] and testing [3,4,65], as well as general surveys of the whole field of SBSE [21,36,48]. This paper does not seek to duplicate these surveys, though some material is repeated from them (with permission), where it is relevant and appropriate. Rather, this paper aims to provide those unfamiliar with SBSE with a tutorial and practical guide. The aim is that, having read this paper, the reader will be able to begin to develop SBSE solutions to a chosen software engineering problem and will be able to collect and analyse the results of the applicati...