Machine Learning Explainability
28 February, 2024 – Virtual
This one-day training course will help prevent your models from being a black box. It will provide you with a toolbox of Machine-Learning explainability techniques that you can use to explain your models to technical and non-technical people alike. You will also learn when, how, and why you should use the different techniques, and their drawbacks.
Looking to upskill your team(s) or organization?
Diego will gladly help you further with custom training solutions.
Get in touchDuration
1 day
Time
09:00 – 17:00
Language
English
Lunch
Included
Certification
No
Level
Professional
What will you learn?
After the training, you will be able to:
Understand the use cases for model explainability (debug, bias detection, right to explanation, etc.)
Understand when model explainability is not enough (correlation vs. causality, fairness)
Categorize the methods into sensitivity vs. impact and explain single predictions (local explainability) vs. multiple predictions (global explainability)
Apply the methods with the provided packages
Explain the inner workings of all methods Articulate the downsides for each method
Evaluate whether a method is appropriate for the business use case
Key takeaways
- Break Down Plots for Additive Attributions
- Partial Dependence Plots
- Individual Conditional Expectation Curves / Ceteris Paribus Plots
- Permutation Feature importance
- Shapley and SHAP Values
Program
This one-day training covers the abovementioned techniques and teaches you how to implement them using Scikit-Learn, Dalex, and Shap.
- Introduction to Machine Learning interpretability/explainability
- Example-based explanations
- The inherent interpretability of ML models
- Ceteris Paribus plots
- Break-down Plots for Additive Attributions
- (Model-specific & Permutation) Feature importance
- Partial dependence plots
- Shapley and SHAP values
Who is it for?
This training is perfect for anyone already working with Machine Learning models and wishing to improve and understand them in order to make better predictions.
Requirements
Proficiency in scikit-learn will be required.
To make the most of this advanced training, knowledge of pipelines, column transformers, linear models, and more complex models (e.g., random forests and gradient boosting), and the core concepts of Machine Learning will be necessary.
Machine learning models are being entrusted to perform data-driven decision-making. However, these models can often be a “black box”, meaning the motivations behind our decision-making can be incomprehensible. In this training, you will learn techniques to interpret your model and explain its decision to technical and non-technical people alike.
What else
should I know?
After registering for this course, you will receive a confirmation email with practical information.
See you soon!
Requirements
Travel and accommodation expenses are not included
All literature and course materials are included in the price.