Timezone: »
Predictive multiplicity occurs when classification models with statistically indistinguishable performances assign conflicting predictions to individual samples. When used for decision-making in applications of consequence (e.g., lending, education, criminal justice), models developed without regard for predictive multiplicity may result in unjustified and arbitrary decisions for specific individuals. We introduce a new metric, called Rashomon Capacity, to measure predictive multiplicity in probabilistic classification. Prior metrics for predictive multiplicity focus on classifiers that output thresholded (i.e., 0-1) predicted classes. In contrast, Rashomon Capacity applies to probabilistic classifiers, capturing more nuanced score variations for individual samples. We provide a rigorous derivation for Rashomon Capacity, argue its intuitive appeal, and demonstrate how to estimate it in practice. We show that Rashomon Capacity yields principled strategies for disclosing conflicting models to stakeholders. Our numerical experiments illustrate how Rashomon Capacity captures predictive multiplicity in various datasets and learning models, including neural networks. The tools introduced in this paper can help data scientists measure and report predictive multiplicity prior to model deployment.
Author Information
Hsiang Hsu (Harvard University)
I am Hsiang Hsu, a Harvard Ph.D. student working with Flavio Calmon, and also a Meta Fellow. My research interests lie in promoting the interpretability of representations, improving privacy and fairness, and understanding prediction uncertainty in machine learning. I believe these are important issues in modern machine learning when trying to deploy the models in practice.
Flavio Calmon (Harvard University)
More from the Same Authors
-
2021 : Who Gets the Benefit of the Doubt? Racial Bias in Machine Learning Algorithms Applied to Secondary School Math Education »
Haewon Jeong · Michael D. Wu · Nilanjana Dasgupta · Muriel Medard · Flavio Calmon -
2022 : Improving Generalization with Physical Equations »
Antoine Wehenkel · Jens Behrmann · Hsiang Hsu · Guillermo Sapiro · Gilles Louppe · Joern-Henrik Jacobsen -
2023 Poster: Adapting Fairness Interventions to Missing Values »
Raymond Feng · Flavio Calmon · Hao Wang -
2023 Poster: Arbitrariness Lies Beyond the Fairness-Accuracy Frontier »
Carol Long · Hsiang Hsu · Wael Alghamdi · Flavio Calmon -
2023 Poster: Aleatoric and Epistemic Discrimination: Fundamental Limits of Fairness Interventions »
Hao Wang · Luxi He · Rui Gao · Flavio Calmon -
2022 Panel: Panel 1C-7: Beyond Adult and… & Uncalibrated Models Can… »
Kailas Vodrahalli · Flavio Calmon -
2022 Poster: Beyond Adult and COMPAS: Fair Multi-Class Prediction via Information Projection »
Wael Alghamdi · Hsiang Hsu · Haewon Jeong · Hao Wang · Peter Michalak · Shahab Asoodeh · Flavio Calmon -
2022 Poster: On the Epistemic Limits of Personalized Prediction »
Lucas Monteiro Paes · Carol Long · Berk Ustun · Flavio Calmon -
2021 Poster: Analyzing the Generalization Capability of SGLD Using Properties of Gaussian Channels »
Hao Wang · Yizhe Huang · Rui Gao · Flavio Calmon -
2019 : Poster Session »
Clement Canonne · Kwang-Sung Jun · Seth Neel · Di Wang · Giuseppe Vietri · Liwei Song · Jonathan Lebensold · Huanyu Zhang · Lovedeep Gondara · Ang Li · FatemehSadat Mireshghallah · Jinshuo Dong · Anand D Sarwate · Antti Koskela · Joonas Jälkö · Matt Kusner · Dingfan Chen · Mi Jung Park · Ashwin Machanavajjhala · Jayashree Kalpathy-Cramer · · Vitaly Feldman · Andrew Tomkins · Hai Phan · Hossein Esfandiari · Mimansa Jaiswal · Mrinank Sharma · Jeff Druce · Casey Meehan · Zhengli Zhao · Hsiang Hsu · Davis Railsback · Abraham Flaxman · · Julius Adebayo · Aleksandra Korolova · Jiaming Xu · Naoise Holohan · Samyadeep Basu · Matthew Joseph · My Thai · Xiaoqian Yang · Ellen Vitercik · Michael Hutchinson · Chenghong Wang · Gregory Yauney · Yuchao Tao · Chao Jin · Si Kai Lee · Audra McMillan · Rauf Izmailov · Jiayi Guo · Siddharth Swaroop · Tribhuvanesh Orekondy · Hadi Esmaeilzadeh · Kevin Procopio · Alkis Polyzotis · Jafar Mohammadi · Nitin Agrawal -
2018 : Poster session: Contributed papers »
Michael Cvitkovic · Arijit Patra · Yunpeng Li · RAHMAN BANYA SAFF SANYA · Guanghua Chi · Benjamin Huynh · Hamed Alemohammad · Simón Ramírez Amaya · Nazmus Saquib · Jade Abbott · Teo de Campos · Viraj Prabhu · Alvaro Riascos · Hafte Abera · praney dubey · Tanushyam Chattopadhyay · Hsiang Hsu · Mayank Jain · Kartikeya Bhardwaj · Gabriel Cadamuro · Bradley Gram-Hansen · Georg Dorffner -
2018 : Posters 1 »
Wei Wei · Flavio Calmon · Travis Dick · Leilani Gilpin · Maroussia Lévesque · Malek Ben Salem · Michael Wang · Jack Fitzsimons · Dimitri Semenovich · Linda Gu · Nathaniel Fruchter -
2017 Poster: Optimized Pre-Processing for Discrimination Prevention »
Flavio Calmon · Dennis Wei · Bhanukiran Vinzamuri · Karthikeyan Natesan Ramamurthy · Kush Varshney