All of the links, references, etc., for my course

Introduction_to_Testing_Machine_Learning_Models_Course.md

Introduction to Testing Machine Learning Models

📗 Google Drive with the code and files needed to follow along

😎 Free Course by @CarlosKidman

• Chapter 1: What is Machine Learning?
• Chapter 2: Build a Machine Learning Model
• Chapter 3: Where do Testers fit in Machine Learning?
• Chapter 4: Adversarial Attacks
• Chapter 5: Behavioral Testing
• Chapter 6: Fair and Responsible AI
• Chapter 7: Machine Learning Models in Production

Chapter 1

What is Machine Learning?

No links

Chapter 2

Build a Machine Learning Model

📗 Google Colab Notebook: Predict House Prices

Download the train.csv and test.csv from the Google Drive

Chapter 3

Where do Testers fit in Machine Learning?

References

• Invaluable Data Science Lessons To Learn From The Failure of Zillow's Flipping Business by Sharan Kumar Ravindran
• 5 Key Data Science Learnings from Zillow's 'iBuying' Failures by Pritish Jadhav

Chapter 4

Adversarial Attacks

📗 Google Colab Notebook: Adversarial Attacks

References

• Turtle classified as rifle YouTube video
• Adversarial Attacks Introduction
• Introduction to Adversarial Machine Learning by floydhub.com

Chapter 5

Behavioral Attacks

📗 Google Colab Notebook: Behavioral Attacks

References

• Behavioral Testing white paper by Ribeiro et al.

Chapter 6

Fair and Responsible AI

References

• A Framework for Understanding Sources of Harm throughout the Machine Learning Life Cycle by Harini Suresh and John Guttag
• Identifying Bias in AI by Alexis Cook (on Kaggle University)
• Machine Learning Fairness: Lessons Learned from Google I/O 2019 (video)
• Responsible AI from TensorFlow

Chapter 7

Machine Learning Models in Production

💡 Takeaway terms: Machine Learning Systems and MLOps

References

• huggingface.co
• distilbert model

I went through this course at Test Automation University. Thanks a lot for making it available.

The wording of the following question in chapter 6 got me confused:
«1. Historical Bias occurs when the state of the world in which the data was generated was flawed»
The correct answer was that this statement is true.

Is the word «flawed» used appropriately here? In particular, should it be applied to «the state of the world»?

I based my answer on the assumption that the cause of historical bias is that the state of the world changes and therefore data sampled from it can become outdated. It seems to me that this has little to do with data or «the state of the world» being flawed at some point in the past.