More Like this

1. Coevolution of cognitive abilities and identity signals in individual recognition systems

   https://doi.org/10.1098/rstb.2019.0467  

   Miller, Sara E.; Sheehan, Michael J.; Reeve, H. Kern (July 2020, Philosophical Transactions of the Royal Society B: Biological Sciences)

   Social interactions are mediated by recognition systems, meaning that the cognitive abilities or phenotypic diversity that facilitate recognition may be common targets of social selection. Recognition occurs when a receiver compares the phenotypes produced by a sender with a template. Coevolution between sender and receiver traits has been empirically reported in multiple species and sensory modalities, though the dynamics and relative exaggeration of traits from senders versus receivers have received little attention. Here, we present a coevolutionary dynamic model that examines the conditions under which senders and receivers should invest effort in facilitating individual recognition. The model predicts coevolution of sender and receiver traits, with the equilibrium investment dependent on the relative costs of signal production versus cognition. In order for recognition to evolve, initial sender and receiver trait values must be above a threshold, suggesting that recognition requires some degree of pre-existing diversity and cognitive abilities. The analysis of selection gradients demonstrates that the strength of selection on sender signals and receiver cognition is strongest when the trait values are furthest from the optima. The model provides new insights into the expected strength and dynamics of selection during the origin and elaboration of individual recognition, an important feature of social cognition in many taxa. This article is part of the theme issue ‘Signal detection theory in recognition systems: from evolving models to experimental tests’. 

   more » « less
2. Signals Among Signals: Prioritizing Nongenetic Associations in Massive Data Sets

   https://doi.org/10.1093/aje/kwz031  

   Manrai, Arjun K; Ioannidis, John P; Patel, Chirag J (March 2019, American Journal of Epidemiology)
3. Reconciling Individual Probability Forecasts✱

   https://doi.org/10.1145/3593013.3593980  

   Roth, Aaron; Tolbert, Alexander; Weinstein, Scott (June 2023, ACM Conference on Fairness Accountability and Transparency)

   Individual probabilities refer to the probabilities of outcomes that are realized only once: the probability that it will rain tomorrow, the probability that Alice will die within the next 12 months, the probability that Bob will be arrested for a violent crime in the next 18 months, etc. Individual probabilities are fundamentally unknowable. Nevertheless, we show that two parties who agree on the data—or on how to sample from a data distribution—cannot agree to disagree on how to model individual probabilities. This is because any two models of individual probabilities that substantially disagree can together be used to empirically falsify and improve at least one of the two models. This can be efficiently iterated in a process of “reconciliation” that results in models that both parties agree are superior to the models they started with, and which themselves (almost) agree on the forecasts of individual probabilities (almost) everywhere. We conclude that although individual probabilities are unknowable, they are contestable via a computationally and data efficient process that must lead to agreement. Thus we cannot find ourselves in a situation in which we have two equally accurate and unimprovable models that disagree substantially in their predictions—providing an answer to what is sometimes called the predictive or model multiplicity problem. 

   more » « less
4. Individual Fairness in Pipelines

   Dwork, C; Ilvento, C; Jagadeesan, M (January 2020, Leibniz international proceedings in informatics)

   Roth, A (Ed.)

   It is well understood that a system built from individually fair components may not itself be individually fair. In this work, we investigate individual fairness under pipeline composition. Pipelines differ from ordinary sequential or repeated composition in that individuals may drop out at any stage, and classification in subsequent stages may depend on the remaining “cohort” of individuals. As an example, a company might hire a team for a new project and at a later point promote the highest performer on the team. Unlike other repeated classification settings, where the degree of unfairness degrades gracefully over multiple fair steps, the degree of unfairness in pipelines can be arbitrary, even in a pipeline with just two stages. Guided by a panoply of real-world examples, we provide a rigorous framework for evaluating different types of fairness guarantees for pipelines. We show that naïve auditing is unable to uncover systematic unfairness and that, in order to ensure fairness, some form of dependence must exist between the design of algorithms at different stages in the pipeline. Finally, we provide constructions that permit flexibility at later stages, meaning that there is no need to lock in the entire pipeline at the time that the early stage is constructed. 

   more » « less
5. Individual Fairness Under Uncertainty

   Zhang, Wenbin; Wang, Zichong; Kim, Juyong; Cheng, Cheng; Oommen, Thomas; Ravikumar, Pradeep; Weiss, Jeremy (October 2023, Proceedings of the 26th European Conference on Artificial Intelligence (ECAI))