Nber Working Paper Series Human Decisions and Machine Predictions Human Decisions and Machine Predictions

By Himabindu Lakkaraju. 2017

Subfields: Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Safety, Robustness & Interpretability

Citations in the corpus

Measurement and Fairness (2019, background): “explicitly consider how to minimize or compensate for historical injustices [11]. This nontrivial measurement task is deeply complex and requires signicant debate about justice, policy, and fairness [8, 13, 15, 37]. 4.2 Value-Added Modeling in Education We now look to the well-known example of value”
Diagnosing Gender Bias in Image Recognition Systems (2020, neutral): “The promise of machine-learning algorithms has been that they would bypass this aspect of human bias, leading to more accurate or equitable results (Cowgill 2018; Gates, Perry, and Zorn 2002; Kleinberg, Ludwig et al. 2018).”
Prediction-Based Decisions and Fairness: A Catalogue of Choices, Assumptions, and Definitions (2018, background): “This problem is sometimes called selective labels [72, 79, 13, 18, 26].”
Fairwashing: the risk of rationalization (2019, background): “, credit scoring (Siddiqi, 2012), predictive justice (Kleinberg et al., 2017) or medical diagnosis (Caruana et al.”
Deep k-Nearest Neighbors: Towards Confident, Interpretable and Robust Deep Learning (2018, background): “, admitting a student to university [101], predicting if a defendant awaiting trial can be released safely [102], modeling consumer credit risk [103].”
Interventions over Predictions: Reframing the Ethical Debate for Actuarial Risk Assessment (2017, background): “they are designed to advise or supplant [46].”
Interpretability via Model Extraction (2017, background): “Recent advances in machine learning have revolutionized our ability to use data to inform critical decisions, such as medical diagnosis [8, 19, 27], bail decisions for defendants [16, 17], and aircra collision avoidance systems [25].”
Proportionally Fair Clustering (2019, background): “Motivated by concerns ranging from the granting of bail in the legal system to the quality of recommender systems, researchers have devoted considerable effort to developing fair algorithms for the canonical supervised learning tasks of classification and regression (Dwork et al., 2012; Kleinberg et”
Using Simpson's Paradox to Discover Interesting Patterns in Behavioral Data (2018, background): “Data mining algorithms have explored behavioral data to test social psychology and decision theories (Kleinberg et al. 2017; Bond et al. 2012) and obtain new insights into factors affecting online behavior.”
Computational social scientist beware: Simpson’s paradox in behavioral data (2017, neutral): “Algorithms have mined behavioral data to validate theories of individual decision making and social interaction [5, 12] and produce new insights into ﬁrst principles of human behavior.”
Fairness in Criminal Justice Risk Assessments: The State of the Art (2017, background): “How actions are affected is a very important, but different, matter (Berk 2017; Kleinberg et al. 2017).”
Characterizing Fairness Over the Set of Good Models Under Selective Labels (2021, background): “A key empirical challenge in domains such as credit lending is that outcomes are not observed for all cases (Lakkaraju et al., 2017; Kleinberg et al., 2018).”
"How do I fool you?": Manipulating User Trust via Misleading Black Box Explanations (2019, background): “We use a dataset on bail outcomes collected from several state courts in the U.S. between 1990-2009 (Lakkaraju et al. 2019).”
Learning Fair Rule Lists (2019, background): “Machine learning models are now becoming more and more common in high stake decision-making systems ( e.g. , credit scoring [56], predictive justice [37] and automatic recruiting [47]).”
Manipulating and Measuring Model Interpretability (2018, methodology): “[37] Jon Kleinberg, Himabindu Lakkaraju, Jure Leskovec, Jens Ludwig, and Sendhil Mullainathan.”