{"product_id":"machine-learning-safety-9789811968136","title":"Machine Learning Safety","description":"\u003cp\u003e\u003c\/p\u003e\u003cblockquote\u003e\n\u003cbr\u003eMachine learning algorithms allow computers to learn without being explicitly programmed, but this raises safety concerns due to its susceptibility to environmental noise and adversarial attacks. This book addresses these concerns by presenting up-to-date techniques for adversarial attacks, formal verification, and adversarial training to improve readers' awareness and equip them with practical skills. \u003c\/blockquote\u003e\u003cp\u003e\u003cstrong\u003eFormat\u003c\/strong\u003e: Hardback\u003cbr\u003e\u003cstrong\u003eLength\u003c\/strong\u003e: 321 pages\u003cbr\u003e\u003cstrong\u003ePublication date\u003c\/strong\u003e: 06 February 2023\u003cbr\u003e\u003cstrong\u003ePublisher\u003c\/strong\u003e: Springer Verlag, Singapore\u003cbr\u003e\u003c\/p\u003e \u003cp\u003e\u003cbr\u003e\u003c\/p\u003e\u003cp\u003e\u003cbr\u003eMachine learning algorithms enable computers to acquire knowledge and skills without being explicitly programmed. Their widespread application spans various domains, including medical diagnostics and fully autonomous vehicles. While this advancement holds immense potential, it also brings forth concerns regarding safety. Machine learning exhibits distinct characteristics that differentiate it from explicitly programmed software systems, making it challenging to predict and assess its behavior.\u003cbr\u003e\u003cbr\u003eThis book delves into the primary safety concerns associated with machine learning, encompassing its susceptibility to environmental noise and adversarial attacks. These vulnerabilities have posed significant barriers to the deployment of machine learning in safety-critical applications. The book presents cutting-edge techniques for adversarial attacks, which are employed to evaluate the vulnerabilities of machine learning models. Formal verification is also utilized to ascertain whether a trained machine learning model is free from vulnerabilities. Adversarial training is employed to enhance the training process and mitigate vulnerabilities.\u003cbr\u003e\u003cbr\u003eThe primary objective of this book is to raise awareness among readers about the potential safety issues related to machine learning models. Furthermore, it equips readers with the latest techniques for addressing these concerns, providing them with a comprehensive understanding of both technical principles and practical skills. By exploring these topics, the book aims to empower individuals and organizations to leverage the benefits of machine learning while ensuring the highest levels of safety and reliability.\u003cbr\u003e\u003c\/p\u003e\u003cp\u003e\u003cbr\u003e\u003cstrong\u003eDimension\u003c\/strong\u003e: 235 x 155 (mm)\u003cbr\u003e\u003cstrong\u003eISBN-13\u003c\/strong\u003e: 9789811968136\u003cbr\u003e \u003cstrong\u003eEdition number\u003c\/strong\u003e: 1st ed. 2023\u003c\/p\u003e","brand":"Xiaowei Huang,Gaojie Jin,Wenjie Ruan","offers":[{"title":"Hardback","offer_id":44235823120634,"sku":"9789811968136","price":54.13,"currency_code":"GBP","in_stock":false}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/0522\/4297\/2845\/products\/1683902887101_book.jpg?v=1684161889","url":"https:\/\/shulphink.com\/products\/machine-learning-safety-9789811968136","provider":"Shulph Ink","version":"1.0","type":"link"}