91精品国产综合久久四虎久久_国产成人午夜高潮毛片_99er视频精品免费观看_2020亚洲熟女在线观看_日本女优人体写真_国内黄色毛片_年轻的老师中文版在线_丰满女邻居做爰_久久久久久精品成人免费图片

 
可靠的機器學習（影印版）
可靠的機器學習（影印版）
Cathy Chen, Niall Richard Murphy, Kranti Parisa, D. Sculley, Todd Underwood 著
出版時間：2023年03月
頁數(shù)：376
“在將基于機器學習的真實系統(tǒng)投入部署之前，你便能從閱讀本書中受益。放心吧，書中的內容來自數(shù)十年間來之不易的經驗。”
——Andrew Moore
Google Cloud AI副總裁兼總經理

無論你是小型創(chuàng)業(yè)公司還是跨國公司的一員，這本實踐用書都為你（數(shù)據科學家、軟件和網站可靠性工程師、產品經理或企業(yè)主）展示了如何在組織內可靠、有效和負責地運行和建立機器學習。你將深入了解其中涉及的方方面面，從如何在生產中進行模型監(jiān)控到如何在產品組織中運營一個完善的模型開發(fā)團隊。
通過將SRE思維應用于機器學習，作為本書作者和工程專業(yè)人士的Cathy Chen、Kranti Parisa、Niall Richard Murphy、D. Sculley、Todd Underwood以及特邀作者向你展示了如何運行高效可靠的機器學習系統(tǒng)。無論你是想增加收入、優(yōu)化決策、解決問題，還是想理解和影響客戶行為，你都將學到如何執(zhí)行日常的機器學習任務，同時保持更廣闊的視野。
本書內容包括：
● 什么是ML：運作方式以及依賴什么
● 用于理解機器學習“環(huán)路”如何工作的概念框架
● 有效的生產如何使機器學習系統(tǒng)易于監(jiān)控、部署和操作
● 為什么機器學習系統(tǒng)使生產故障排除更加困難，以及如何進行相應的補償
● 機器學習、產品和生產團隊如何有效溝通
  1. Foreword
  2. Preface
  3. 1. Introduction
  4. The ML Lifecycle
  5. Lessons from the Loop
  6. 2. Data Management Principles
  7. Data as Liability
  8. The Data Sensitivity of ML Pipelines
  9. Phases of Data
  10. Data Reliability
  11. Data Integrity
  12. Conclusion
  13. 3. Basic Introduction to Models
  14. What Is a Model?
  15. A Basic Model Creation Workflow
  16. Model Architecture Versus Model Definition Versus Trained Model
  17. Where Are the Vulnerabilities?
  18. Infrastructure and Pipelines
  19. A Set of Useful Questions to Ask About Any Model
  20. An Example ML System
  21. Conclusion
  22. 4. Feature and Training Data
  23. Features
  24. Labels
  25. Human-Generated Labels
  26. Metadata
  27. Data Privacy and Fairness
  28. Conclusion
  29. 5. Evaluating Model Validity and Quality
  30. Evaluating Model Validity
  31. Evaluating Model Quality
  32. Operationalizing Verification and Evaluation
  33. Conclusion
  34. 6. Fairness, Privacy, and Ethical ML Systems
  35. Fairness (a.k.a. Fighting Bias)
  36. Privacy
  37. Responsible AI
  38. Responsible AI Along the ML Pipeline
  39. Conclusion
  40. 7. Training Systems
  41. Requirements
  42. Basic Training System Implementation
  43. General Reliability Principles
  44. Common Training Reliability Problems
  45. Structural Reliability
  46. Conclusion
  47. 8. Serving
  48. Key Questions for Model Serving
  49. Model Serving Architectures
  50. Model API Design
  51. Serving for Accuracy or Resilience?
  52. Scaling
  53. Disaster Recovery
  54. Ethics and Fairness Considerations
  55. Conclusion
  56. 9. Monitoring and Observability for Models
  57. What Is Production Monitoring and Why Do It?
  58. Problems with ML Production Monitoring
  59. Best Practices for ML Model Monitoring
  60. Conclusion
  61. 10. Continuous ML
  62. Anatomy of a Continuous ML System
  63. Observations About Continuous ML Systems
  64. Continuous Organizations
  65. Rethinking Noncontinuous ML Systems
  66. Conclusion
  67. 11. Incident Response
  68. Incident Management Basics
  69. Anatomy of an ML-Centric Outage
  70. Terminology Reminder: Model
  71. Story Time
  72. ML Incident Management Principles
  73. Special Topics
  74. Conclusion
  75. 12. How Product and ML Interact
  76. Different Types of Products
  77. Agile ML?
  78. ML Product Development Phases
  79. Build Versus Buy
  80. Sample YarnIt Store Features Powered by ML
  81. Conclusion
  82. 13. Integrating ML into Your Organization
  83. Chapter Assumptions
  84. Significant Organizational Risks
  85. Implementation Models
  86. Organizational Design and Incentives
  87. Conclusion
  88. 14. Practical ML Org Implementation Examples
  89. Scenario 1: A New Centralized ML Team
  90. Scenario 2: Decentralized ML Infrastructure and Expertise
  91. Scenario 3: Hybrid with Centralized Infrastructure/Decentralized Modeling
  92. Conclusion
  93. 15. Case Studies: MLOps in Practice
  94. 1. Accommodating Privacy and Data Retention Policies in ML Pipelines
  95. 2. Continuous ML Model Impacting Traffic
  96. 3. Steel Inspection
  97. 4. NLP MLOps: Profiling and Staging Load Test
  98. 5. Ad Click Prediction: Databases Versus Reality
  99. 6. Testing and Measuring Dependencies in ML Workflow
  100. Index
書名：可靠的機器學習（影印版）
國內出版社：東南大學出版社
出版時間：2023年03月
頁數(shù)：376
書號：978-7-5766-0552-5
原版書書名：Reliable Machine Learning
原版書出版商：O'Reilly Media
Cathy Chen
 
Cathy Chen曾在Google擔任技術項目經理、產品經理和工程經理。
 
 
Niall Richard Murphy
 
Niall Richard Murphy是Google網站可靠性工程組織里曾經和現(xiàn)任的成員，他們的職責是關注和維護Google的生產系統(tǒng)。
 
 
Kranti Parisa
 
Kranti Parisa是Dialpad的副總裁兼產品工程主管。
 
 
D. Sculley
 
D. Sculley是Kaggle的首席執(zhí)行官和Google第三方機器學習生態(tài)系統(tǒng)的總經理。
 
 
Todd Underwood
 
Todd Underwood是Google的高級主管以及機器學習SRE的創(chuàng)始人。
 
 
The insect on the cover of Reliable Machine Learning is the honeypot ant (Myrmecocystus mimicus). Honeypot ants are found in southwest North America and parts of Mexico.
Similar to other ants, honeypot ant colonies consist of a variety of worker ants who scavenge food from flowers, fruit, and other insects. What is most notable about honeypot ants is how they store food. The repletes—one type of worker ant in the colony—grow large abdomens that they use to store the liquid they scavenge. During times when food supply is low, the repletes regurgitate liquid for the rest of the colony to eat. Repletes have a hard time moving around because of the size of their abdomen, so they are often found hanging from the roof of their nest.
購買選項
定價：119.00元
書號：978-7-5766-0552-5
出版社：東南大學出版社