MAX-PLANCK-INSTITUT FÜR WISSENSCHAFTSGESCHICHTE Max Planck Institute for the History of Science

Event

May 23, 2023
Explainable AI in the Digital Humanities

Our speakers are Oliver Eberle, Machine Learning Researcher at TU Berlin, and Hassan El-Hajj. They will talk on "Explainable AI in the Digital Humanities."

Explainable AI (XAI) has become an active subfield of machine learning that aims to make predictions of modern, highly non-linear deep models more understandable, transparent, and actionable. To achieve this, explanation techniques, such as layer-wise relevance propagation (LRP), visualize the influential features of a model's prediction. Thus, XAI is crucial for building more robust and fair machine learning models and is especially needed when models are used for scientific insight discovery. In this talk, I will give a brief introduction to XAI and highlight the utility of explanations in the context of digital humanities.

Organizer(s)
Address
MPIWG, Boltzmannstraße 22, 14195 Berlin, Germany
Room
Room 265 & Online
Contact and Registration

For further information and registration, please contact Kim Pham: kpham@mpiwg-berlin.mpg.de

About This Series

Brown Bag Lunch is a meeting of researchers at the MPIWG who use or want to learn more about digital research methods, broadly encompassed by the term Digital Humanities. In the Brown Bag Lunch meetings, researchers can discuss tools, share ideas and experiences (good and bad), and learn from each other. Each session explores a new topic; workshops are usually interactive, and we often invite external speakers.

2023-05-23T12:00:00SAVE IN I-CAL 2023-05-23 12:00:00 2023-05-23 13:30:00 Explainable AI in the Digital Humanities Our speakers are Oliver Eberle, Machine Learning Researcher at TU Berlin, and Hassan El-Hajj. They will talk on "Explainable AI in the Digital Humanities." Explainable AI (XAI) has become an active subfield of machine learning that aims to make predictions of modern, highly non-linear deep models more understandable, transparent, and actionable. To achieve this, explanation techniques, such as layer-wise relevance propagation (LRP), visualize the influential features of a model's prediction. Thus, XAI is crucial for building more robust and fair machine learning models and is especially needed when models are used for scientific insight discovery. In this talk, I will give a brief introduction to XAI and highlight the utility of explanations in the context of digital humanities. MPIWG, Boltzmannstraße 22, 14195 Berlin, Germany Room 265 & Online Kim Pham Kim Pham Europe/Berlin public