Machine Learning Safety

The history of robotics and artificial intelligence in many ways is also the history of humanity’s attempts to control such technologies. From the Golem of Prague to the military robots of modernity, the debate continues as to what degree of independence such entities should have and how to make sure that they do not turn on us, its inventors. Numerous recent advancements in all aspects of research, development and deployment of intelligent systems are well publicized but safety and security issues related to AI are rarely addressed. This book is proposed to mitigate this fundamental problem. It is comprised of chapters from leading AI Safety researchers addressing different aspects of the AI control problem as it relates to the development of safe and secure artificial intelligence. The book is the first edited volume dedicated to addressing challenges of constructing safe and secure advanced machine intelligence. The chapters vary in length and technical content from broad interest opinion essays to highly formalized algorithmic approaches to specific problems. All chapters are self-contained and could be read in any order or skipped without a loss of comprehension.

Machine learning algorithms allow computers to learn without being explicitly programmed. Their application is now spreading to highly sophisticated tasks across multiple domains, such as medical diagnostics or fully autonomous vehicles. While this development holds great potential, it also raises new safety concerns, as machine learning has many specificities that make its behaviour prediction and assessment very different from that for explicitly programmed software systems. This book addresses the main safety concerns with regard to machine learning, including its susceptibility to environmental noise and adversarial attacks. Such vulnerabilities have become a major roadblock to the deployment of machine learning in safety-critical applications. The book presents up-to-date techniques for adversarial attacks, which are used to assess the vulnerabilities of machine learning models; formal verification, which is used to determine if a trained machine learning model is free of vulnerabilities; and adversarial training, which is used to enhance the training process and reduce vulnerabilities. The book aims to improve readers’ awareness of the potential safety issues regarding machine learning models. In addition, it includes up-to-date techniques for dealing with these issues, equipping readers with not only technical knowledge but also hands-on practical skills.

Can machine learning techniques solve our computer security problems and finally put an end to the cat-and-mouse game between attackers and defenders? Or is this hope merely hype? Now you can dive into the science and answer this question for yourself. With this practical guide, you’ll explore ways to apply machine learning to security issues such as intrusion detection, malware classification, and network analysis. Machine learning and security specialists Clarence Chio and David Freeman provide a framework for discussing the marriage of these two fields, as well as a toolkit of machine-learning algorithms that you can apply to an array of security problems. This book is ideal for security engineers and data scientists alike. Learn how machine learning has contributed to the success of modern spam filters Quickly detect anomalies, including breaches, fraud, and impending system failure Conduct malware analysis by extracting useful information from computer binaries Uncover attackers within the network by finding patterns inside datasets Examine how attackers exploit consumer-facing websites and app functionality Translate your machine learning algorithms from the lab to production Understand the threat attackers pose to machine learning solutions

Introduces Machine Learning Techniques and Tools and Provides Guidance on How to Implement Machine Learning Into Chemical Safety and Health-related Model Development There is a growing interest in the application of machine learning algorithms in chemical safety and health-related model development, with applications in areas including property and toxicity prediction, consequence prediction, and fault detection. This book is the first to review the current status of machine learning implementation in chemical safety and health research and to provide guidance for implementing machine learning techniques and algorithms into chemical safety and health research. Written by an international team of authors and edited by renowned experts in the areas of process safety and occupational and environmental health, sample topics covered within the work include: An introduction to the fundamentals of machine learning, including regression, classification and cross-validation, and an overview of software and tools Detailed reviews of various applications in the areas of chemical safety and health, including flammability prediction, consequence prediction, asset integrity management, predictive nanotoxicity and environmental exposure assessment, and more Perspective on the possible future development of this field Machine Learning in Chemical Safety and Health serves as an essential guide on both the fundamentals and applications of machine learning for industry professionals and researchers in the fields of process safety, chemical safety, occupational and environmental health, and industrial hygiene.

Nearly all our safety data collection and reporting systems are backwardlooking: incident reports; dashboards; compliance monitoring systems; and so on. This book shows how we can use safety data in a forward-looking, predictive sense. Predictive Safety Analytics: Reducing Risk through Modeling and Machine Learning contains real use cases where organizations have reduced incidents by employing predictive analytics to foresee and mitigate future risks. It discusses how Predictive Safety Analytics is an opportunity to break through the plateau problem where safety rate improvements have stagnated in many organizations. The book presents how the use of data, coupled with advanced analytical techniques, including machine learning, has become a proven and successful innovation. Emphasis is placed on how the book can “meet you where you are” by illuminating a path to get there, starting with simple data the organization likely already has. Highlights of the book are the real examples and case studies that will assist in generating thoughts and ideas for what might work for individual readers and how they can adapt the information to their particular situations. This book is written for professionals and researchers in system reliability, risk and safety assessment, quality control, operational managers in selected industries, data scientists, and ML engineers. Students taking courses in these areas will also find this book of interest to them.

As the advancement of technology continues, cyber security continues to play a significant role in today’s world. With society becoming more dependent on the internet, new opportunities for virtual attacks can lead to the exposure of critical information. Machine and deep learning techniques to prevent this exposure of information are being applied to address mounting concerns in computer security. The Handbook of Research on Machine and Deep Learning Applications for Cyber Security is a pivotal reference source that provides vital research on the application of machine learning techniques for network security research. While highlighting topics such as web security, malware detection, and secure information sharing, this publication explores recent research findings in the area of electronic security as well as challenges and countermeasures in cyber security research. It is ideally designed for software engineers, IT specialists, cybersecurity analysts, industrial experts, academicians, researchers, and post-graduate students.

Can we make machines that think and act like humans or other natural intelligent agents? The answer to this question depends on how we see ourselves and how we see the machines in question. Classical AI and cognitive science had claimed that cognition is computation, and can thus be reproduced on other computing machines, possibly surpassing the abilities of human intelligence. This consensus has now come under threat and the agenda for the philosophy and theory of AI must be set anew, re-defining the relation between AI and Cognitive Science. We can re-claim the original vision of general AI from the technical AI disciplines; we can reject classical cognitive science and replace it with a new theory (e.g. embodied); or we can try to find new ways to approach AI, for example from neuroscience or from systems theory. To do this, we must go back to the basic questions on computing, cognition and ethics for AI. The 30 papers in this volume provide cutting-edge work from leading researchers that define where we stand and where we should go from here.

This book constitutes the proceedings of the Workshops held in conjunction with SAFECOMP 2020, 39th International Conference on Computer Safety, Reliability and Security, Lisbon, Portugal, September 2020. The 26 regular papers included in this volume were carefully reviewed and selected from 45 submissions; the book also contains one invited paper. The workshops included in this volume are: DECSoS 2020: 15th Workshop on Dependable Smart Embedded and Cyber-Physical Systems and Systems-of-Systems. DepDevOps 2020: First International Workshop on Dependable Development-Operation Continuum Methods for Dependable Cyber-Physical Systems. USDAI 2020: First International Workshop on Underpinnings for Safe Distributed AI. WAISE 2020: Third International Workshop on Artificial Intelligence Safety Engineering. The workshops were held virtually due to the COVID-19 pandemic.