Distributed Machine Learning: Algorithms, Theory and Practice Chinese edition? Latest electronic PDF download:

 

In the era of artificial intelligence and big data, distributed machine learning is the mainstream solution to solve the most challenging problems! The purpose of this book is to introduce the current situation of distributed machine learning, analyze the core technical issues, and discuss the future development direction of this field.

Machine learning core team of Microsoft Research Asia! Academician E Weinan and Professor Zhou Zhihua wrote the recommendation preface!

The purpose of this book is to introduce the current situation of distributed machine learning, analyze the core technical issues, and discuss the future development direction of this field.

The book consists of 12 chapters. Chapter 1 is the introduction, showing you the field of distributed machine learning. Chapter 2 introduces the basics of machine learning. Chapters 3 through 8 are the core of the book, providing a detailed introduction to the distributed machine learning framework and its functional modules. Chapter 3 gives an overview of the whole distributed machine learning framework, while chapters 4 to 8 introduce data and model partitioning module, single machine optimization module, communication module, data and model aggregation module respectively. The following three chapters are the summary and sublimation of the previous content. Chapter 9 introduces various distributed machine learning algorithms combined with different options in distributed machine learning frameworks, Chapter 10 discusses the theoretical properties of these algorithms, and Chapter 11 introduces several mainstream distributed machine learning systems (including Spark MLlib iterative MapReduce system, Multiverso parameter server system, TensorFlow data flow system). Chapter 12 is the conclusion of the book. After a brief summary of the book, it focuses on the future development direction of distributed machine learning.

This book is based on the research achievements and practical experience of the machine learning research team of Microsoft Research Asia. It can be used as a reference for graduate students engaged in the research of distributed machine learning, and also as a reference book for artificial intelligence practitioners to choose algorithms and design systems.

There are many machine learning books on the market, but few dedicated books on distributed machine learning. This book is a boon for readers who want to learn and understand distributed machine learning.

 

 

directory

Introduction Chapter 1 Introduction / 1 1.1 Artificial Intelligence and its Rapid Development / 2 1.2 Large-scale and Distributed machine learning / 4 1.3 Book arrangement / 6 References / 7 Chapter 2 Fundamentals of machine learning / 9 2.1 Basic Concepts of machine learning / 10 2.2 Basic Flow of machine learning / 13 2.3 Commonly used Loss functions / 16 2.4 Commonly used Machine learning models / 18 2.5 Commonly used optimization methods / 32 2.6 Machine learning theory / 33 2.7 Summary / 36 References / 36 Chapter 3 Distributed Machine Learning Frameworks / 41 3.1 Challenges of Big Data and Big Model / 42 3.2 Basic Flow of Distributed machine learning / 44 3.3 Data and Model Partitioning module / 46 3.4 Stand-alone optimization module / 48 3.5 Communication module / 48 3.6 Data and Model aggregation module / 53 3.7 Distributed machine learning theory / 4.1 Basic Overview / 62 4.2 First-order deterministic algorithms / 67 4.3 Second-order deterministic algorithms / 75 4.4 Dual methods / 78 5.1 Basic Stochastic optimization algorithm / 86 5.2 Improvement of stochastic optimization algorithm / 96 5.3 Non-convex stochastic optimization algorithm / 101 5.4 Summary / 109 References / 109 Chapter 6 Data and Model Parallel / 113 6.1 Basic Overview / 114 6.2 Computational Parallel mode / 117 6.3 Data parallel mode / 119 6.4 Model parallel mode / 123 6.5 Summary / 133 References / 133 Chapter 7 Communication mechanism / 135 7.1 Basic Overview / 136 7.2 Content of Communication / 137 7.3 Topology of Communication / 139 7.4 Pace of communication / 145 7.5 Frequency of communication / 150 7.6 Summary / 156 References / 156 Chapter 8 Data and Model Aggregation / 159 8.1 Basic Overview / 160 8.2 Model sum-based Aggregation methods / 160 8.4 Summary / 174 References / 174 Chapter 9 Distributed Machine Learning Algorithms / 177 9.1 Basic Overview / 178 9.2 Synchronous algorithms / 179 9.3 Asynchronous algorithms / 187 9.4 Comparison and Fusion of Synchronous and asynchronous / 199 9.5 Model Parallel Algorithms / 203 9.6 Summary / 205 References / 205 Chapter 10 distributed machine learning Theory / 209 10.1 Basic Overview / 210 10.2 Convergence analysis / 210 10.2.3 Synchronous and Asynchronous / 215 10.3 Acceleration ratio Analysis / 217 10.4 Generalization Analysis / 221 10.5 Summary / 226 References / 226 Chapter 11 distributed machine learning Systems / 229 11.1 Basic Overview / 230 11.2 Distributed Machine Learning System Based on IMR / 231 11.3 Distributed Machine Learning System Based on Parameter Server / 236 11.4 Distributed Machine Learning System Based on Data Stream / 241 11.5 Actual Comparison / 248 11.6 Summary / 252 References / 252 Chapter 12 Epilogue / 255 Index / 260


 

Click on the download