Shapley values is an attribution system devised by economist Lloyd Shapley based on the Cooperative Game theory. We can use it to describe a machine learning model's output. In game theory, the Shapley value is a notion used to determine each player's contribution in a coalition or cooperative game. It has lately gained attention as an effective way to explain ML learning model predictions. 

Shapley value is a popular strategy derived from cooperative game theory with beneficial qualities. For example, it is commonly used to explain why someone has been rejected for a loan by ML models in the lending business. This article will provide the reader with the best resources for learning about Shapley Values in Machine Learning.

Machine Learning Modeling Pipelines in Production - Coursera

In this course of Machine Learning Engineering for Production Specialization:

• you will build models for different serving environments, 
• use tools and techniques to effectively manage your modelling resources, and 
• use analytics tools and performance metrics to address model fairness, explainability issues, and bottlenecks.

Machine learning engineering for production combines the basic ideas of machine learning with the functional knowledge of modern software development and engineering roles to help you develop skills that are ready for production.

Interpretable Machine Learning - By Christoph Molnar

This book discusses making machine learning models and their decisions more interpretable. You will learn about generic model-independent methods for analysing black box models, such as feature importance and accumulated local effects, and explain individual predictions using Shapley values and LIME. All interpretation approaches are well explained and critically debated. It will teach you how to choose and appropriately implement the best interpretation method for your machine-learning project.

Understanding Machine Learning From Theory to Algorithms

This textbook's purpose is to present machine learning and its algorithmic paradigms systematically. This book gives a comprehensive theoretical analysis of the fundamental concepts underlying machine learning and the mathematical derivations that translate these principles into practical algorithms. Following a review of the field's fundamentals, the book covers many vital issues not covered in earlier textbooks. 

It includes a discussion of the computational complexity of learning, significant algorithmic paradigms and emerging theoretical concepts like the PAC-Bayes approach and compression-based bounds. The text makes the fundamentals and techniques of machine learning accessible to students. It is designed for an advanced undergraduate or beginning graduate course.

Machine Learning and Artificial Intelligence - MLIS 2020

MLIS 2020 was the most recent in a series of yearly conferences designed to facilitate the exchange of information regarding the most recent scientific and technological advancements in machine learning and intelligent systems. Additionally, the annual conference builds connections among the scientific community in linked study fields. The publication features 53 articles chosen from more than 160 submissions and presented at MLIS 2020. Data mining, image processing, neural networks, human health, natural language processing, video processing, computational intelligence, expert systems, human-computer interaction, deep learning, and robotics are some topics covered. The book will interest individuals working on the subject because it provides an overview of current research and advancements in machine learning and artificial intelligence.