16

Aug

Teaser image to How to "Improve" Prediction Using Behavior Modification

Lecture

How to "Improve" Prediction Using Behavior Modification

Galit Shmueli, National Tsing Hua University, Taiwan

   16.08.2024

   2:30 pm - 3:30 pm

   LMU Munich, Lecture Hall S 002, Ground Floor, Schellingstr. 3, 80799 Munich

Galit Shmueli will explore how internet platforms leverage behavioral data to predict user behavior, both for internal purposes and for their business clients, such as advertisers, insurers, and governments.

Achieving high predictive accuracy is essential, and data scientists are continually developing advanced algorithms and utilizing more extensive and diverse datasets to improve it. However, these platforms can also boost prediction accuracy by subtly steering users' behavior to match predicted outcomes, using behavior modification techniques. This approach, often involving reinforcement learning algorithms, is not widely addressed in the machine learning and statistics literature. To thoroughly examine this strategy, the authors introduce causal reasoning, specifically Pearl's causal do(.) operator, into predictive modeling. They dissect the prediction error that arises when behavior modification is employed and discuss its implications for prediction accuracy, platform users, and business clients. While behavior modification can make predictions seem more accurate, it may not translate to effective real-world applications and could have unintended negative consequences for users.

Organized by:

Institute of AI in Management LMU Munich


Related

Link to Foundational Methods for Foundation Models for Scientific Machine Learning

Munich AI Lectures  •  09.07.2025  •  Plenarsaal of the Bavarian Academy of Sciences and Humanities (BAdW), Alfons-Goppel-Straße 11, 80539 Munich

Foundational Methods for Foundation Models for Scientific Machine Learning

The Munich AI Lectures on March 26, 2025, will feature Michael Mahoney and cover foundational methods for SciML foundation models.


Link to The Best of CVPR 2025 Series

Meetup  •  09.07.2025  •  Online

The Best of CVPR 2025 Series

Voxel51’s “Best of CVPR” event features top 2025 computer vision research, including Cecilia Curreli presenting on human motion prediction.


Link to Festival of the Future 2025

Festival  •  03.07.2025 - 06.07.2025  •  Deutsches Museum, Munich

Festival of the Future 2025

Festival of the Future 2025 takes place from July 3 to 6, 2025, covering AI, sustainability, and space, with Björn Ommer as a speaker.


Link to Bayesian Insights on Polarization and Disagreement

Lecture  •  26.06.2025  •  CAS Seestraße 13 80802 Munich

Bayesian Insights on Polarization and Disagreement

As part of the CAS Research Focus, Leah Henderson explores societal polarization using Bayesian models, offering insights on handling disagreements.