Book Image

Hands-On Explainable AI (XAI) with Python

By : Denis Rothman

Book Image

Hands-On Explainable AI (XAI) with Python

By: Denis Rothman

Overview of this book

Effectively translating AI insights to business stakeholders requires careful planning, design, and visualization choices. Describing the problem, the model, and the relationships among variables and their findings are often subtle, surprising, and technically complex. Hands-On Explainable AI (XAI) with Python will see you work with specific hands-on machine learning Python projects that are strategically arranged to enhance your grasp on AI results analysis. You will be building models, interpreting results with visualizations, and integrating XAI reporting tools and different applications. You will build XAI solutions in Python, TensorFlow 2, Google Cloud’s XAI platform, Google Colaboratory, and other frameworks to open up the black box of machine learning models. The book will introduce you to several open-source XAI tools for Python that can be used throughout the machine learning project life cycle. You will learn how to explore machine learning model results, review key influencing variables and variable relationships, detect and handle bias and ethics issues, and integrate predictions using Python along with supporting the visualization of machine learning models into user explainable interfaces. By the end of this AI book, you will possess an in-depth understanding of the core concepts of XAI.

Preface

Who this book is for

What this book covers

To get the most out of this book

Explaining Artificial Intelligence with Python

Explaining Artificial Intelligence with Python

Defining explainable AI

Designing and extracting

The XAI medical diagnosis timeline

Further reading

Free Chapter

White Box XAI for AI Bias and Ethics

White Box XAI for AI Bias and Ethics

Moral AI bias in self-driving cars

Standard explanation of autopilot decision trees

XAI applied to an autopilot decision tree

Using XAI and ethics to control a decision tree

Further reading

Explaining Machine Learning with Facets

Explaining Machine Learning with Facets

Getting started with Facets

Facets Overview

Sorting the Facets statistics overview

Further reading

Microsoft Azure Machine Learning Model Interpretability with SHAP

Microsoft Azure Machine Learning Model Interpretability with SHAP

Introduction to SHAP

Getting started with SHAP

Linear models and logistic regression

Further reading

Building an Explainable AI Solution from Scratch

Building an Explainable AI Solution from Scratch

Moral, ethical, and legal perspectives

The U.S. census data problem

The machine learning perspective

WIT applied to a transformed dataset

Further reading

AI Fairness with Google's What-If Tool (WIT)

AI Fairness with Google's What-If Tool (WIT)

Interpretability and explainability from an ethical AI perspective

Getting started with WIT

Creating a DNN model

Creating a SHAP explainer

Model outputs and SHAP values

The WIT datapoint explorer and editor

Further reading

A Python Client for Explainable AI Chatbots

A Python Client for Explainable AI Chatbots

The Python client for Dialogflow

Enhancing the Google Dialogflow Python client

A CUI XAI dialog using Google Dialogflow

Further reading

Local Interpretable Model-Agnostic Explanations (LIME)

Local Interpretable Model-Agnostic Explanations (LIME)

Introducing LIME

Getting started with LIME

An experimental AutoML module

Interpreting the scores

Training the model and making predictions

The LIME explainer

Further reading

The Counterfactual Explanations Method

The Counterfactual Explanations Method

The counterfactual explanations method

The choice of distance functions

The architecture of the deep learning model

Further reading

Contrastive XAI

Contrastive XAI

The contrastive explanations method

Getting started with the CEM applied to MNIST

Defining and training the CNN model

Defining and training the autoencoder

Pertinent negatives

Further reading

Anchors XAI

Anchors AI explanations

Anchor explanations for ImageNet

Further reading

Cognitive XAI

Cognitive rule-based explanations

A cognitive approach to vectorizers

Human cognitive input for the CEM

Further reading

Answers to the Questions

Chapter 1, Explaining Artificial Intelligence with Python

Chapter 2, White Box XAI for AI Bias and Ethics

Chapter 3, Explaining Machine Learning with Facets

Chapter 4, Microsoft Azure Machine Learning Model Interpretability with SHAP

Chapter 5, Building an Explainable AI Solution from Scratch

Chapter 6, AI Fairness with Google's What-If Tool (WIT)

Chapter 7, A Python Client for Explainable AI Chatbots

Chapter 8, Local Interpretable Model-Agnostic Explanations (LIME)

Chapter 9, The Counterfactual Explanations Method

Chapter 10, Contrastive XAI

Chapter 11, Anchors XAI

Chapter 12, Cognitive XAI

Other Books You May Enjoy

Other Books You May Enjoy

Index

Customer Reviews

5 star

0

4 star

0

3 star

0

2 star

0

1 star

0

Anchor explanations for ImageNet

In this section, we will build a Python Alibi program that will produce anchors for images. Alibi is a library that contains several XAI resources.

We will use images from ImageNet to run the explainer.

We will build the program in the following order:

Installing Alibi and importing the modules
Loading an InceptionV3 model
Downloading an image to explain
Processing the image and making predictions
Creating the anchor image explainer and displaying the visual explanations

Let's first install Alibi and import the modules.

Installing Alibi and importing the modules

To get started, open the Image_XAI_Anchor.ipynb notebook for this chapter. This notebook is in the chapter directory of this book.

We will first install Alibi as follows:

# @title Install Alibi
try:
  import alibi
except:
  !pip install alibi

If Alibi is installed, the program will continue. However, when Colaboratory...