/

Ripple-Down Rules The Alternative to Machine Learning

Compton Paul Kang Byeong Ho
Heftet / 2021 / Engelsk

Produktdetaljer

ISBN
9780367644321
Publisert
2021-05-31
Utgiver
Vendor
Chapman & Hall/CRC
Vekt
660 gr
Høyde
234 mm
Bredde
156 mm
Aldersnivå
U, P, 05, 06
Språk
Product language
Engelsk
Format
Product format
Heftet
Antall sider
196

Forfatter
Compton Paul
Kang Byeong Ho

Biographical note

Paul Compton initially studied philosophy before majoring in physics. He spent 20 years as a biophysicist at the Garvan Institute of Medical Research, and then 20 years in Computer Science and Engineering at the University of New South Wales, where he was head of school for 12 years and is now an emeritus professor.

Byeong Ho Kang majored in mathematics in Korea, followed by a PhD on Ripple-Down Rules at the University of New South Wales and the algorithm he developed is the basis of most industry RDR applications. He is a professor, with a research focus on applications, and head of the ICT discipline at the University of Tasmania."

Ripple-Down Rules The Alternative to Machine Learning

Compton Paul Kang Byeong Ho
Heftet / 2021 / Engelsk
Compton Paul Kang Byeong Ho
Heftet / 2021 / Engelsk
Nettpris:
768,-
Laveste pris siste 30 dager: 768,-
Levering 10-30 dager
Klikk og hent

<p>"In this era of deep learning, where is our deeper understanding of AI? The answer is, here, in this book. Compton and Kang's ideas are a "must-read" for anyone working with AI. Based on very many examples of real-world applications, they show us a better way to use AI. If your AI models are confusing to understand and hard to maintain, then this book is for you."</p><p>-- <strong><em>Tim Menzies, Professor, North Carolina State University</em></strong></p>

<p>"In this era of deep learning, where is our deeper understanding of AI? The answer is, here, in this book. Compton and Kang's ideas are a "must-read" for anyone working with AI. Based on very many examples of real-world applications, they show us a better way to use AI. If your AI models are confusing to understand and hard to maintain, then this book is for you."</p><p>-- <strong><em>Tim Menzies, Professor, North Carolina State University</em></strong></p>

Machine learning algorithms hold extraordinary promise, but the reality is that their success depends entirely on the suitability of the data available. This book is about Ripple-Down Rules (RDR), an alternative manual technique for rapidly building AI systems. With a human in the loop, RDR is much better able to deal with the limitations of data.Ripple-Down Rules: The Alternative to Machine Learning starts by reviewing the problems with data quality and the problems with conventional approaches to incorporating expert human knowledge into AI systems. It suggests that problems with knowledge acquisition arise because of mistaken philosophical assumptions about knowledge. It argues people never really explain how they reach a conclusion, rather they justify their conclusion by differentiating between cases in a context. RDR is based on this more situated understanding of knowledge. The central features of a RDR approach are explained, and detailed worked examples are presented for different types of RDR, based on freely available software developed for this book. The examples ensure developers have a clear idea of the simple yet counter-intuitive RDR algorithms to easily build their own RDR systems. It has been proven in industrial applications that it takes only a minute or two per rule to build RDR systems with perhaps thousands of rules. The industrial uses of RDR have ranged from medical diagnosis through data cleansing to chatbots in cars. RDR can be used on its own or to improve the performance of machine learning or other methods.
Les mer
Machine learning algorithms hold out extraordinary promise, but the reality is that their success depends entirely on the suitability of the data available. This book is about Ripple-Down Rules an alternative manual technique for rapidly building AI systems. With a human in the loop, RDR is much better able to deal with the limitations of the data.
Les mer
PrefaceAcknowledgements1 Problems with Machine Learning and Knowledge Acquisition1.1 Introduction1.2 Machine Learning1.3 Knowledge Acquisition2 Philosophical issues in knowledge acquisition3 Ripple-Down Rule Overview3.1 Case-driven knowledge acquisition3.2 Order of cases processed3.3 Linked Production Rules3.4 Adding rules3.5 Assertions and retractions3.6 Formulae in conclusion4 Introduction to Excel_RDR5 Single Classification Example5.1 Repetition in an SCRDR knowledge base5.2 SCRDR evaluation and machine learning comparison5.3 Summary6 Multiple classification example6.1 Introduction to Multiple Classification Ripple-Down Rules (MCRDR)6.2 Excel_MCRDR example6.3 Discussion: MCRDR for single classification6.4 Actual Multiple classification with MCRDR6.5 Discussion6.6 Summary7 General Ripple-Down Rules (GRDR)7.1 Background7.2 Key Features of GRDR7.3 Excel_GRDR demo7.4 Discussion: GRDR, MCRDR and SCRDR8 Implementation and Deployment of RDR-based systems8.1 Validation8.2 The role of the user/expert8.3 Cornerstone Cases8.4 Explanation_8.5 Implementation Issues8.6 Information system interfaces9 RDR and Machine learning9.1 Suitable datasets9.2 Human experience versus statistics.9.3 Unbalanced Data9.4 Prudence9.5 RDR-based machine learning methods9.6 Machine learning combined with RDR knowledge acquisition9.7 Machine learning supporting RDR9.8 Summary_ Appendix 1 - Industrial Applications of RDRA1.1 PEIRS (1991-1995)A1.2 Pacific Knowledge SystemsA1.3 IvisA1.4 Erudine Pty LtdA1.5 Ripple-Down Rules at IBMA1.6 YAWLA1.7 MedscopeA1.8 SeegeneA1.9 IPMSA1.10 Tapacross Appendix 2 - Research-demonstrated ApplicationsA2.1 RDR WrappersA2.2 Text-processing, natural language processing and information retrievalA2.3 Conversational agents and help desksA2.4 RDR for operator and parameter selectionA2.5 Anomaly and event detectionA2.6 RDR for image and video processing ReferencesIndex
Les mer
"In this era of deep learning, where is our deeper understanding of AI? The answer is, here, in this book. Compton and Kang's ideas are a "must-read" for anyone working with AI. Based on very many examples of real-world applications, they show us a better way to use AI. If your AI models are confusing to understand and hard to maintain, then this book is for you."-- Tim Menzies, Professor, North Carolina State University
Les mer

Relaterte produkter

<p>"In this era of deep learning, where is our deeper understanding of AI? The answer is, here, in this book. Compton and Kang's ideas are a "must-read" for anyone working with AI. Based on very many examples of real-world applications, they show us a better way to use AI. If your AI models are confusing to understand and hard to maintain, then this book is for you."</p><p>-- <strong><em>Tim Menzies, Professor, North Carolina State University</em></strong></p>

<p>"In this era of deep learning, where is our deeper understanding of AI? The answer is, here, in this book. Compton and Kang's ideas are a "must-read" for anyone working with AI. Based on very many examples of real-world applications, they show us a better way to use AI. If your AI models are confusing to understand and hard to maintain, then this book is for you."</p><p>-- <strong><em>Tim Menzies, Professor, North Carolina State University</em></strong></p>

Machine learning algorithms hold extraordinary promise, but the reality is that their success depends entirely on the suitability of the data available. This book is about Ripple-Down Rules (RDR), an alternative manual technique for rapidly building AI systems. With a human in the loop, RDR is much better able to deal with the limitations of data.Ripple-Down Rules: The Alternative to Machine Learning starts by reviewing the problems with data quality and the problems with conventional approaches to incorporating expert human knowledge into AI systems. It suggests that problems with knowledge acquisition arise because of mistaken philosophical assumptions about knowledge. It argues people never really explain how they reach a conclusion, rather they justify their conclusion by differentiating between cases in a context. RDR is based on this more situated understanding of knowledge. The central features of a RDR approach are explained, and detailed worked examples are presented for different types of RDR, based on freely available software developed for this book. The examples ensure developers have a clear idea of the simple yet counter-intuitive RDR algorithms to easily build their own RDR systems. It has been proven in industrial applications that it takes only a minute or two per rule to build RDR systems with perhaps thousands of rules. The industrial uses of RDR have ranged from medical diagnosis through data cleansing to chatbots in cars. RDR can be used on its own or to improve the performance of machine learning or other methods.
Les mer
Machine learning algorithms hold out extraordinary promise, but the reality is that their success depends entirely on the suitability of the data available. This book is about Ripple-Down Rules an alternative manual technique for rapidly building AI systems. With a human in the loop, RDR is much better able to deal with the limitations of the data.
Les mer
PrefaceAcknowledgements1 Problems with Machine Learning and Knowledge Acquisition1.1 Introduction1.2 Machine Learning1.3 Knowledge Acquisition2 Philosophical issues in knowledge acquisition3 Ripple-Down Rule Overview3.1 Case-driven knowledge acquisition3.2 Order of cases processed3.3 Linked Production Rules3.4 Adding rules3.5 Assertions and retractions3.6 Formulae in conclusion4 Introduction to Excel_RDR5 Single Classification Example5.1 Repetition in an SCRDR knowledge base5.2 SCRDR evaluation and machine learning comparison5.3 Summary6 Multiple classification example6.1 Introduction to Multiple Classification Ripple-Down Rules (MCRDR)6.2 Excel_MCRDR example6.3 Discussion: MCRDR for single classification6.4 Actual Multiple classification with MCRDR6.5 Discussion6.6 Summary7 General Ripple-Down Rules (GRDR)7.1 Background7.2 Key Features of GRDR7.3 Excel_GRDR demo7.4 Discussion: GRDR, MCRDR and SCRDR8 Implementation and Deployment of RDR-based systems8.1 Validation8.2 The role of the user/expert8.3 Cornerstone Cases8.4 Explanation_8.5 Implementation Issues8.6 Information system interfaces9 RDR and Machine learning9.1 Suitable datasets9.2 Human experience versus statistics.9.3 Unbalanced Data9.4 Prudence9.5 RDR-based machine learning methods9.6 Machine learning combined with RDR knowledge acquisition9.7 Machine learning supporting RDR9.8 Summary_ Appendix 1 - Industrial Applications of RDRA1.1 PEIRS (1991-1995)A1.2 Pacific Knowledge SystemsA1.3 IvisA1.4 Erudine Pty LtdA1.5 Ripple-Down Rules at IBMA1.6 YAWLA1.7 MedscopeA1.8 SeegeneA1.9 IPMSA1.10 Tapacross Appendix 2 - Research-demonstrated ApplicationsA2.1 RDR WrappersA2.2 Text-processing, natural language processing and information retrievalA2.3 Conversational agents and help desksA2.4 RDR for operator and parameter selectionA2.5 Anomaly and event detectionA2.6 RDR for image and video processing ReferencesIndex
Les mer
"In this era of deep learning, where is our deeper understanding of AI? The answer is, here, in this book. Compton and Kang's ideas are a "must-read" for anyone working with AI. Based on very many examples of real-world applications, they show us a better way to use AI. If your AI models are confusing to understand and hard to maintain, then this book is for you."-- Tim Menzies, Professor, North Carolina State University
Les mer

Produktdetaljer

ISBN
9780367644321
Publisert
2021-05-31
Utgiver
Vendor
Chapman & Hall/CRC
Vekt
660 gr
Høyde
234 mm
Bredde
156 mm
Aldersnivå
U, P, 05, 06
Språk
Product language
Engelsk
Format
Product format
Heftet
Antall sider
196

Forfatter
Compton Paul
Kang Byeong Ho

Biographical note

Paul Compton initially studied philosophy before majoring in physics. He spent 20 years as a biophysicist at the Garvan Institute of Medical Research, and then 20 years in Computer Science and Engineering at the University of New South Wales, where he was head of school for 12 years and is now an emeritus professor.

Byeong Ho Kang majored in mathematics in Korea, followed by a PhD on Ripple-Down Rules at the University of New South Wales and the algorithm he developed is the basis of most industry RDR applications. He is a professor, with a research focus on applications, and head of the ICT discipline at the University of Tasmania."

Relaterte produkter