You're reading from Interpretable Machine Learning with Python Build explainable, fair, and robust high-performance models with hands-on, real-world examples

Product type Paperback

Published in Oct 2023

Publisher Packt

ISBN-13 9781803235424

Length 606 pages

Edition 2nd Edition

Languages

Python

Tools

BERT

Concepts

Machine Learning

Author (1):

Serg Masís

View More author details

Table of Contents (17) Chapters

Preface

1. Interpretation, Interpretability, and Explainability; and Why Does It All Matter?

2. Key Concepts of Interpretability FREE CHAPTER

3. Interpretation Challenges

4. Global Model-Agnostic Interpretation Methods

5. Local Model-Agnostic Interpretation Methods

6. Anchors and Counterfactual Explanations

7. Visualizing Convolutional Neural Networks

8. Interpreting NLP Transformers

9. Interpretation Methods for Multivariate Forecasting and Sensitivity Analysis

10. Feature Selection and Engineering for Interpretability

11. Bias Mitigation and Causal Inference Methods

12. Monotonic Constraints and Model Tuning for Interpretability

13. Adversarial Robustness

14. What’s Next for Machine Learning Interpretability?

15. Other Books You May Enjoy

16. Index

The mission

Imagine you are an analyst for a national health ministry, and there’s a Cardiovascular Diseases (CVDs) epidemic. The minister has made it a priority to reverse the growth and reduce the caseload to a 20-year low. To this end, a task force has been created to find clues in the data to ascertain the following:

What risk factors can be addressed.
If future cases can be predicted, interpret predictions on a case-by-case basis.

You are part of this task force!

Details about CVD

Before we dive into the data, we must gather some important details about CVD in order to do the following:

Understand the problem’s context and relevance.
Extract domain knowledge information that can inform our data analysis and model interpretation.
Relate an expert-informed background to a dataset’s features.

CVDs are a group of disorders, the most common of which is coronary heart disease (also known as Ischaemic Heart Disease). According to the World Health Organization, CVD is the leading cause of death globally, killing close to 18 million people annually. Coronary heart disease and strokes (which are, for the most part, a byproduct of CVD) are the most significant contributors to that. It is estimated that 80% of CVD is made up of modifiable risk factors. In other words, some of the preventable factors that cause CVD include the following:

Poor diet
Smoking and alcohol consumption habits
Obesity
Lack of physical activity
Poor sleep

Also, many of the risk factors are non-modifiable and, therefore, known to be unavoidable, including the following:

Genetic predisposition
Old age
Male (varies with age)

We won’t go into more domain-specific details about CVD because it is not required to make sense of the example. However, it can’t be stressed enough how central domain knowledge is to model interpretation. So, if this example was your job and many lives depended on your analysis, it would be advisable to read the latest scientific research on the subject and consult with domain experts to inform your interpretations.

You're reading from Interpretable Machine Learning with Python Build explainable, fair, and robust high-performance models with hands-on, real-world examples

Table of Contents (17) Chapters

The mission

Details about CVD

Authors (1)

Personalised recommendations for you