Skip to main content
main-content

Über dieses Buch

This book is written for software product teams that use AI to add intelligent models to their products or are planning to use it. As AI adoption grows, it is becoming important that all AI driven products can demonstrate they are not introducing any bias to the AI-based decisions they are making, as well as reducing any pre-existing bias or discrimination.

The responsibility to ensure that the AI models are ethical and make responsible decisions does not lie with the data scientists alone. The product owners and the business analysts are as important in ensuring bias-free AI as the data scientists on the team. This book addresses the part that these roles play in building a fair, explainable and accountable model, along with ensuring model and data privacy. Each chapter covers the fundamentals for the topic and then goes deep into the subject matter – providing the details that enable the business analysts and the data scientists to implement these fundamentals.

AI research is one of the most active and growing areas of computer science and statistics. This book includes an overview of the many techniques that draw from the research or are created by combining different research outputs. Some of the techniques from relevant and popular libraries are covered, but deliberately not drawn very heavily from as they are already well documented, and new research is likely to replace some of it.

Inhaltsverzeichnis

Frontmatter

Chapter 1. Introduction

Abstract
The machine ethics were mostly a topic for science fiction before the twenty-first century. The easy access and low cost of data storage and computing capabilities have meant that the use of machine-driven intelligence has increased significantly in the last two decades and the topic is not one for entertainment or a theoretical one anymore. The machine learning algorithms or AI now impacts our lives in innumerable ways – making decisions that can have material impact on the lives of the individuals affected by these decisions. Our experience has shown us that all roles within a product team contribute to building an effective AI product, and, in this book, we have tried to cover the content from the point of view of all the roles and not just the data engineers and the data scientists.
Sray Agarwal, Shashin Mishra

Chapter 2. Fairness and Proxy Features

Abstract
In this chapter we will look at how different fairness metrics can help us understand whether our data is fair or not. We will also look at what are proxy features, how they can impact the fairness in the data and the different techniques we can use to detect proxy features in our dataset.
Sray Agarwal, Shashin Mishra

Chapter 3. Bias in Data

Abstract
In this chapter, we are going to discuss how unintended bias can get introduced in the models through the data and how the product owner and the SMEs can work with the data scientists at the definition stage to identify the biases that can get introduced and make decisions on how to reduce or eliminate them.
Sray Agarwal, Shashin Mishra

Chapter 4. Explainability

Abstract
This chapters covers AI explainability from three lenses: explaining features, explaining models and creating explainable models. For each of these approaches, it goes deep into the detail of how you can utilize them, the scenarios that make one option preferred over the others and multiple ways of applying them.
Sray Agarwal, Shashin Mishra

Chapter 5. Remove Bias from ML Model

Abstract
In this chapter, we will be covering the techniques to overcome the bias in the data. We have picked up the techniques that are model agnostic and can be used with most of the machine learning algorithms. These techniques offer a lot of scope for optimization and the business analysts in the team will have to take the lead in understanding and deciding the best approach for the optimization that can be achieved.
Sray Agarwal, Shashin Mishra

Chapter 6. Remove Bias from ML Output

Abstract
In the previous chapter, we saw two techniques that help us address the bias either before the model training by adding weights to the records (reweighting) or by adding a step in the modelling process by calculating the residuals (which requires additional model training). Both of these techniques come in when you still haven’t trained your model and allow you to build a model that is fair from grounds up. However, these techniques do not help us if we already have models in production that are doing predictions today that may have bias learnt from the historical data used for training them. In this chapter, we will discuss how you can address the bias after a model has made its predictions.
Sray Agarwal, Shashin Mishra

Chapter 7. Accountability in AI

Abstract
In this chapter, we will look at the concept of model drift and the types of drifts – data and concept. For each of these, we will then explain the various methods that can be used by the business analysts and the data scientists alike to determine if a drift is happening and ways to determine the right method to apply given a specific problem.
Sray Agarwal, Shashin Mishra

Chapter 8. Data and Model Privacy

Abstract
There are different techniques that can be used to improve the privacy in the data and the model. The improved privacy does not just help protect the sensitive parameters in the data but also help reduce the bias. This chapter explains why you should be considering data and model privacy as an integral part of your responsible AI journey and how you can apply it for the problem you are working on.
Sray Agarwal, Shashin Mishra

Chapter 9. Conclusion

Abstract
AI has massive impact on the lives it touches, and the responsibility to build RAI lies with everyone building intelligent products. This chapter provides some practical tools that you can use to integrate responsible AI in your data science lifecycle and discusses how AI has an ethical impact beyond the parameters discussed in the book and the approach that can help continuous development of responsible AI.
Sray Agarwal, Shashin Mishra
Weitere Informationen

Premium Partner