Administrative information

Title	Explainable AI for end-users
Duration	60 mins
Module	B-opt
Lesson Type	Lecture
Focus	Ethical - Trustworthy AI
Topic	General Explainable AI

Keywords

Explainable AI,

Learning Goals

Get a deeper understanding of how XAI can be communicated to end users

Expected Preparation

Learning Events to be Completed Before

None.

Obligatory for Students

Dalex, a package combining various XAI techniques, and offering a tutorial
Analysis of SHAP for visual images Bento, V., Kohler, M., Diaz, P., Mendoza, L., & Pacheco, M. A. (2021). Improving deep learning performance by using Explainable Artificial Intelligence (XAI) approaches. Discover Artificial Intelligence, 1(1), 1-11.
Attacking discrimination in ML
Visual intro to ML

Optional for Students

SHAP explained
SHAP values Credit modeling prediction
Rudin, C. (2019). Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 1(5), 206-215.

References and background for students:

None.

Recommended for Teachers

None.

Lesson Materials

Slides

The materials of this learning event are available under CC BY-NC-SA 4.0.

Instructions for Teachers

Practical laboratory that provides learners with a well understood data set and requires them to apply a machine learning algorithm for classification. Learners will then have to select a post-hoc explainability technique (ICE, DeepLIFT, LIME, SHAP).

Students will have to apply a relevant Explainable AI technique (or methodology) on their specific project/model, and report why their approach is suited and which (ethical) problems it addresses.

Trustworthy AI is a wider concept that just applying (post hoc) XAI techniques. For Trustworthy AI, the primary model should properly be understood (before resorting to XAI tools) (non post-hoc) approach of "explainability" is also be to visualize (explain) the effect of model thresholds on e.g. fairness.

[[Review status::I think this LE can be deleted. There is a lot of overlap with LE156/LE157 (the practicals) and the other two lectures on this topic, LE154, LE158). [OurenKuiper]

This LE is a "lecture", but the description talks about a " practical laboratory"...

!!! Many important fields are empty (e.g. goals, keywords, lesson material, keywords); outline is missing!!!| ]]

More information

Click here for an overview of all lesson plans of the master human centred AI

Please visit the home page of the consortium HCAIM

Acknowledgements

The Human-Centered AI Masters programme was co-financed by the Connecting Europe Facility of the European Union Under Grant №CEF-TC-2020-1 Digital Skills 2020-EU-IA-0068.

The materials of this learning event are available under CC BY-NC-SA 4.0

The HCAIM consortium consists of three excellence centres, three SMEs and four Universities