book

Applied Machine Learning Explainability Techniques

Name: Applied Machine Learning Explainability Techniques
Author: Aditya Bhattacharya
ISBN: 9781803246154

by Aditya Bhattacharya

July 2022

Beginner to intermediate

306 pages

7h 3m

English

Packt Publishing

Read now

Unlock full access

Applied Machine Learning Explainability Techniques
ContributorsAbout the authorAbout the reviewers
Preface
Who this book is forWhat this book coversTo get the most out of this bookDownload the example code filesDownload the color imagesConventions usedGet in touchShare Your Thoughts
Section 1 – Conceptual Exposure
Chapter 1: Foundational Concepts of Explainability Techniques
Introduction to XAIUnderstanding the key termsConsequences of poor predictionsSummarizing the need for model explainabilityDefining explanation methods and approachesDimensions of explainabilityAddressing key questions of explainabilityUnderstanding different types of explanation methodsUnderstanding the accuracy interpretability trade-offEvaluating the quality of explainability methods Criteria for good explainable ML systemsAuxiliary criteria of XAI for ML systemsTaxonomy of evaluation levels for explainable ML systemsSummaryReferences
Chapter 2: Model Explainability Methods
Technical requirements Types of model explainability methodsKnowledge extraction methods EDAResult visualization methodsUsing comparison analysisUsing Surrogate Explainer methodsInfluence-based methodsFeature importanceSensitivity analysisPDPsLRPRepresentation-based explanationVAMsExample-based methodsCFEs in structured dataCFEs in unstructured dataSummaryReferences
Chapter 3: Data-Centric Approaches
Technical requirements Introduction to data-centric XAIAnalyzing data volumeAnalyzing data consistencyAnalyzing data purityThorough data analysis and profiling processThe need for data analysis and profiling processesData analysis as a precautionary stepBuilding robust data profilesMonitoring and anticipating driftsDetecting driftsSelection of statistical measuresChecking adversarial robustnessImpact of adversarial attacksMethods to increase adversarial robustnessEvaluating adversarial robustnessMeasuring data forecastability Estimating data forecastabilitySummaryReferences
Section 2 – Practical Problem Solving
Chapter 4: LIME for Model Interpretability
Technical requirements Intuitive understanding of LIMELearning interpretable data representationsMaintaining a balance in the fidelity-interpretability trade-offSearching for local explorationsWhat makes LIME a good model explainer?SP-LIMEA practical example of using LIME for classification problemsPotential pitfallsSummaryReferences
Chapter 5: Practical Exposure to Using LIME in ML
Technical requirements Using LIME on tabular dataSetting up LIMEDiscussion about the datasetDiscussions about the modelApplication of LIMEExplaining image classifiers with LIMESetting up the required Python modulesUsing a pre-trained TensorFlow model as our black-box modelApplication of LIME Image ExplainersUsing LIME on text dataInstalling the required Python modulesDiscussions about the dataset used for training the modelDiscussions about the text classification model Applying LIME Text ExplainersLIME for production-level systemsSummaryReferences
Chapter 6: Model Interpretability Using SHAP
Technical requirements An intuitive understanding of the SHAP and Shapley valuesIntroduction to SHAP and Shapley valuesWhat are Shapley values?Shapley values in MLThe SHAP frameworkModel explainability approaches using SHAPVisualizations in SHAPExplainers in SHAPUsing SHAP to explain regression modelsSetting up SHAPInspecting the datasetTraining the modelApplication of SHAPAdvantages and limitations of SHAPAdvantages LimitationsSummaryReferences

Chapter 7: Practical Exposure to Using SHAP in ML
Technical requirements Applying TreeExplainers to tree ensemble modelsInstalling the required Python modulesDiscussion about the datasetTraining the modelApplication of TreeExplainer in SHAPExplaining deep learning models using DeepExplainer and GradientExplainerGradientExplainerDiscussion on the dataset used for training the modelUsing a pre-trained CNN model for this exampleApplication of GradientExplainer in SHAPExploring DeepExplainersApplication of DeepExplainer in SHAPModel-agnostic explainability using KernelExplainerApplication of KernelExplainer in SHAPExploring LinearExplainer in SHAPApplication of LinearExplainer in SHAPExplaining transformers using SHAP Explaining transformer-based sentiment analysis modelsExplaining a multi-class prediction transformer model using SHAPExplaining zero-shot learning models using SHAPSummaryReferences
Chapter 8: Human-Friendly Explanations with TCAV
Technical requirements Understanding TCAV intuitivelyWhat is TCAV?Explaining with abstract conceptsGoals of TCAVApproach of TCAVExploring the practical applications of TCAVGetting startedAbout the dataDiscussions about the deep learning model usedModel explainability using TCAVAdvantages and limitationsAdvantagesLimitationsPotential applications of concept-based explanationsSummaryReferences
Chapter 9: Other Popular XAI Frameworks
Technical requirements DALEXSetting up DALEX for model explainabilityDiscussions about the datasetTraining the modelModel explainability using DALEXModel-level explanationsPrediction-level explanationsEvaluating model fairnessInteractive dashboards using ARENAExplainerdashboardSetting up ExplainerdashboardModel explainability with ExplainerdashboardInterpretMLSupported explanation methodsSetting up InterpretMLDiscussions about the datasetTraining the modelExplainability with InterpretMLALIBISetting up ALIBIDiscussion about the datasetTraining the modelModel explainability with ALIBIDiCECFE methods supported in DiCEModel explainability with DiCEELI5Setting up ELI5Model explainability using ELI5H2O AutoML explainersExplainability with H2O explainersQuick comparison guideSummaryReferences
Section 3 –Taking XAI to the Next Level
Chapter 10: XAI Industry Best Practices
Open challenges of XAIGuidelines for designing explainable ML systemsAdopting a data-first approach for explainabilityEmphasizing IML for explainabilityEmphasizing prescriptive insights for explainabilitySummaryReferences
Chapter 11: End User-Centered Artificial Intelligence
User-centered XAI/ML systemsDifferent aspects of end user-centric XAIRapid XAI prototyping using EUCAEfforts toward increasing user acceptance of AI/ML systems using XAIProviding a delightful UXSummaryReferences
Why subscribe?
Other Books You May EnjoyPackt is searching for authors like youShare Your Thoughts

Overview

"Applied Machine Learning Explainability Techniques" is a comprehensive guide to understanding and implementing XAI (Explainable AI). By exploring explainability frameworks such as LIME and SHAP, this book will help you design transparent and trustworthy machine learning systems that promote AI adoption in industrial and research scenarios.

What this Book will help me do

Understand different methods for model explainability and evaluation.
Learn to explain outcomes of structured and unstructured data models.
Gain hands-on experience with tools such as LIME, SHAP, and others.
Master best practices for implementing XAI solutions in the industry.
Design user-centric AI systems to bridge the gap between AI and users.

Author(s)

Aditya Bhattacharya is a seasoned data scientist and AI researcher specializing in explainable machine learning techniques. With significant expertise in AI/ML systems design and implementation, Aditya brings a blend of academic and practical insights. Their approachable style makes complex subjects accessible to learners of various experience levels.

Who is it for?

This book is designed for professionals in machine learning, such as data scientists, AI researchers, ML engineers, and software architects. Ideal for those with basic experience in Python and ML/DL concepts, it caters to individuals aiming to apply XAI effectively for real-world challenges and improve model transparency.

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

Publisher Resources

ISBN: 9781803246154

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills