scispace - formally typeset
Search or ask a question

Showing papers in "Communications of The ACM in 2019"


Journal ArticleDOI
TL;DR: In this paper, the authors provide a survey covering existing techniques to increase interpretability of machine learning models and discuss crucial issues that the community should consider in future work such as designing user-friendly explanations and developing comprehensive evaluation metrics to further push forward the area of interpretable machine learning.
Abstract: Interpretable machine learning tackles the important problem that humans cannot understand the behaviors of complex machine learning models and how these models arrive at a particular decision. Although many approaches have been proposed, a comprehensive understanding of the achievements and challenges is still lacking. We provide a survey covering existing techniques to increase the interpretability of machine learning models. We also discuss crucial issues that the community should consider in future work such as designing user-friendly explanations and developing comprehensive evaluation metrics to further push forward the area of interpretable machine learning.

759 citations


Journal ArticleDOI
TL;DR: Innovations like domain-specific hardware, enhanced security, open instruction sets, and agile chip development will lead the way.
Abstract: Innovations like domain-specific hardware, enhanced security, open instruction sets, and agile chip development will lead the way.

434 citations


Journal ArticleDOI
TL;DR: The kind of causal inference seen in natural human thought can be "algorithmitized" to help produce human-level machine intelligence.
Abstract: The kind of causal inference seen in natural human thought can be "algorithmitized" to help produce human-level machine intelligence.

415 citations



Journal ArticleDOI
TL;DR: This research presents a meta-modelling system that automates the very labor-intensive and therefore time-heavy and therefore expensive and expensive process of manually fixing programming mistakes.
Abstract: Automated program repair can relieve programmers from the burden of manually fixing the ever-increasing number of programming mistakes.

213 citations



Journal ArticleDOI
TL;DR: In this article, the authors rigorously prove the device-independent security of an entanglement-based protocol building on Ekert's original proposal for quantum key distribution, using techniques from the classical theory of pseudo-randomness to achieve a new quantitative understanding of the nonlocal nature of quantum correlations.
Abstract: Quantum cryptography promises levels of security that are impossible to attain in a classical world. Can this security be guaranteed to classical users of a quantum protocol, who may not even trust the quantum devices used to implement the protocol?This central question dates back to the early 1990s when the challenge of achieving Device-Independent Quantum Key Distribution (DIQKD) was first formulated. We answer the challenge by rigorously proving the device-independent security of an entanglement-based protocol building on Ekert's original proposal for quantum key distribution. The proof of security builds on techniques from the classical theory of pseudo-randomness to achieve a new quantitative understanding of the non-local nature of quantum correlations.

181 citations


Journal ArticleDOI
TL;DR: Metamorphic testing can test untestable software, detecting fatal errors in autonomous vehicles' onboard computer systems.
Abstract: Metamorphic testing can test untestable software, detecting fatal errors in autonomous vehicles' onboard computer systems.

159 citations


Journal ArticleDOI
TL;DR: In this paper, the behavior of complex AI algorithms, especially in mission-critical settings, must be made intelligible to the user, and they must be trusted to make decisions intelligibly.
Abstract: To trust the behavior of complex AI algorithms, especially in mission-critical settings, they must be made intelligible.

145 citations


Journal ArticleDOI
TL;DR: Noise pollution is not merely an annoyance but an important problem with broad societal effects that apply to a significant portion of the population, and effective noise mitigation is in the public interest, with the promise of health, economic, and quality-of-life benefits.
Abstract: SONYC integrates sensors, machine listening, data analytics, and citizen science to address noise pollution in New York City.

136 citations


Journal ArticleDOI
TL;DR: Key lessons for designing static analyses tools deployed to find bugs in hundreds of millions of lines of code are learned.
Abstract: Key lessons for designing static analyses tools deployed to find bugs in hundreds of millions of lines of code.

Journal ArticleDOI
TL;DR: Retracing the pivotal privacy and security-related events and ensuing issues from the past year and identifying the key players and issues to watch out for.
Abstract: Retracing the pivotal privacy and security-related events and ensuing issues from the past year.

Journal ArticleDOI
TL;DR: Envisioning computing education that both teaches and empowers is proposed, with a focus on K-12 education.
Abstract: Envisioning computing education that both teaches and empowers.

Journal ArticleDOI
TL;DR: In this paper, a simple, expressive logic based on relations is used to describe designs and automate their analysis, and the authors propose a method to automate the analysis of such designs.
Abstract: Exploiting a simple, expressive logic based on relations to describe designs and automate their analysis.

Journal ArticleDOI
TL;DR: A Harvard-based pilot program integrates class sessions on ethical reasoning into courses throughout its computer science curriculum.
Abstract: A Harvard-based pilot program integrates class sessions on ethical reasoning into courses throughout its computer science curriculum.

Journal ArticleDOI
TL;DR: Considering how block-based programming environments and tools might be used at the introductory level and beyond and beyond.
Abstract: Considering how block-based programming environments and tools might be used at the introductory level and beyond.

Journal ArticleDOI
TL;DR: Separation logic is a key development in formal reasoning about programs, opening up new lines of attack on longstanding problems.
Abstract: Separation logic is a key development in formal reasoning about programs, opening up new lines of attack on longstanding problems.

Journal ArticleDOI
TL;DR: This work design, implement, and evaluate DeepXplore, the first white-box framework for systematically testing real-world DL systems, and introduces neuron coverage for measuring the parts of a DL system exercised by test inputs and demonstrates how finding inputs for DL systems that both trigger many differential behaviors and achieve high neuron coverage can be represented as a joint optimization problem and solved efficiently using gradient-based search techniques.
Abstract: Deep learning (DL) systems are increasingly deployed in safety- and security-critical domains such as self-driving cars and malware detection, where the correctness and predictability of a system's behavior for corner case inputs are of great importance. Existing DL testing depends heavily on manually labeled data and therefore often fails to expose erroneous behaviors for rare inputs. We design, implement, and evaluate DeepXplore, the first white-box framework for systematically testing real-world DL systems. First, we introduce neuron coverage for measuring the parts of a DL system exercised by test inputs. Next, we leverage multiple DL systems with similar functionality as cross-referencing oracles to avoid manual checking. Finally, we demonstrate how finding inputs for DL systems that both trigger many differential behaviors and achieve high neuron coverage can be represented as a joint optimization problem and solved efficiently using gradient-based search techniques. DeepXplore efficiently finds thousands of incorrect corner case behaviors (e.g., self-driving cars crashing into guard rails and malware masquerading as benign software) in state-of-the-art DL models with thousands of neurons trained on five popular datasets such as ImageNet and Udacity self-driving challenge data. For all tested DL models, on average, DeepXplore generated one test input demonstrating incorrect behavior within one second while running only on a commodity laptop. We further show that the test inputs generated by DeepXplore can also be used to retrain the corresponding DL model to improve the model's accuracy by up to 3%.

Journal ArticleDOI
TL;DR: Computer and information scientists join forces with other fields to help solve societal and environmental challenges facing humanity, in pursuit of a sustainable future.
Abstract: Computer and information scientists join forces with other fields to help solve societal and environmental challenges facing humanity, in pursuit of a sustainable future.

Journal ArticleDOI
Maurice Herlihy1
TL;DR: The roots of blockchain technologies are deeply interwoven in distributed computing, and the future of this technology is likely to be driven by the combination of smart contracts and distributed systems.
Abstract: The roots of blockchain technologies are deeply interwoven in distributed computing.

Journal ArticleDOI
TL;DR: In the worst-case analysis of algorithms, the overall performance of an algorithm is summarized by its worst performance on any input as mentioned in this paper, and this approach has countless success stories, but there are also important computational problems such as linear programming, clustering, online caching, and neural network training where the worstcase analysis framework does not provide any helpful advice on how to solve the problem.
Abstract: In the worst-case analysis of algorithms, the overall performance of an algorithm is summarized by its worst performance on any input. This approach has countless success stories, but there are also important computational problems --- like linear programming, clustering, online caching, and neural network training --- where the worst-case analysis framework does not provide any helpful advice on how to solve the problem. This article covers a number of modeling methods for going beyond worst-case analysis and articulating which inputs are the most relevant.

Journal ArticleDOI
TL;DR: In this paper, the authors trace the tangled web of unsolicited and undesired email and possible strategies for its demise, and propose several strategies for their demise, such as email filtering.
Abstract: Tracing the tangled web of unsolicited and undesired email and possible strategies for its demise.

Journal ArticleDOI
TL;DR: Quantum systems will significantly affect the field of cyber security research and will be a major factor in the development of next generation cyber security systems.
Abstract: Quantum systems will significantly affect the field of cyber security research.

Journal ArticleDOI
TL;DR: Diffusion speed and scale depend on all kinds of information, not just which users have the most or fewest connections.
Abstract: Diffusion speed and scale depend on all kinds of information, not just which users have the most or fewest connections.

Journal ArticleDOI
Russ Cox1
TL;DR: Software reuse is finally here, and its benefits should not be understated, but the authors’ve accepted this transformation without completely thinking through the potential consequences.
Abstract: Software reuse is finally here, and its benefits should not be understated, but we’ve accepted this transformation without completely thinking through the potential consequences. The Copay and Equi...

Journal ArticleDOI
TL;DR: In this paper, the abstract ideas behind Kalman filtering are presented at a level accessible to anyone with a basic knowledge of probability theory and calculus, and then they can be applied to the particular problem of state estimation in linear systems.
Abstract: Kalman filtering is a classic state estimation technique used in application areas such as signal processing and autonomous control of vehicles. It is now being used to solve problems in computer systems such as controlling the voltage and frequency of processors. Although there are many presentations of Kalman filtering in the literature, they usually deal with particular systems like autonomous robots or linear systems with Gaussian noise, which makes it difficult to understand the general principles behind Kalman filtering. In this paper, we first present the abstract ideas behind Kalman filtering at a level accessible to anyone with a basic knowledge of probability theory and calculus, and then show how these concepts can be applied to the particular problem of state estimation in linear systems. This separation of concepts from applications should make it easier to understand Kalman filtering and to apply it to other problems in computer systems.

Journal ArticleDOI
TL;DR: Considering the expression "computational thinking" as an entry point to understand why the fundamental contribution of computing to science is the shift from solving problems to having problems so...
Abstract: Considering the expression "computational thinking" as an entry point to understand why the fundamental contribution of computing to science is the shift from solving problems to having problems solved.

Journal ArticleDOI
TL;DR: Programmable software-defined solid-state drives can move computing functions closer to storage by automating the very labor-intensive and therefore time-heavy process of designing and manufacturing these drives.
Abstract: Programmable software-defined solid-state drives can move computing functions closer to storage.

Journal ArticleDOI
TL;DR: Markov logic can be used as a general framework for joining logical and statistical AI.
Abstract: Markov logic can be used as a general framework for joining logical and statistical AI.

Journal ArticleDOI
TL;DR: An approach that allows analysis across multiple versions of QUIC to understand how code changes impact protocol effectiveness and identifies performance issues related to window sizes, re-ordered packets, and multiplexing large number of small objects is developed.
Abstract: Google's Quick UDP Internet Connections (QUIC) protocol, which implements TCP-like properties at the application layer atop a UDP transport, is now used by the vast majority of Chrome clients accessing Google properties but has no formal state machine specification, limited analysis, and ad-hoc evaluations based on snapshots of the protocol implementation in a small number of environ-merits. Further frustrating attempts to evaluate QUIC is the fact that the protocol is under rapid development, with extensive rewriting of the protocol occurring over the scale of months, making individual studies of the protocol obsolete before publication.Given this unique scenario, there is a need for alternative techniques for understanding and evaluating QUIC when compared with previous transport-layer protocols. First, we develop an approach that allows us to conduct analysis across multiple versions of QUIC to understand how code changes impact protocol effectiveness. Next, we instrument the source code to infer QUIC's state machine from execution traces. With this model, we run QUIC in a large number of environments that include desktop and mobile, wired and wireless environments and use the state machine to understand differences in transport-and application-layer performance across multiple versions of QUIC and in different environments. QUIC generally outperforms TCP, but we also identified performance issues related to window sizes, re-ordered packets, and multiplexing large number of small objects; further, we identify that QUIC's performance diminishes on mobile devices and over cellular networks.