Abstract
Sound principles of statistical inference dictate that uncertainty shapes learning. In this work, we revisit the question of learning in volatile environments, in which both the first and second-order statistics of observations dynamically evolve over time. We propose a new model, the volatile Kalman filter (VKF), which is based on a tractable state-space model of uncertainty and extends the Kalman filter algorithm to volatile environments. The proposed model is algorithmically simple and encompasses the Kalman filter as a special case. Specifically, in addition to the error-correcting rule of Kalman filter for learning observations, the VKF learns volatility according to a second error-correcting rule. These dual updates echo and contextualize classical psychological models of learning, in particular hybrid accounts of Pearce-Hall and Rescorla-Wagner. At the computational level, compared with existing models, the VKF gives up some flexibility in the generative model to enable a more faithful approximation to exact inference. When fit to empirical data, the VKF is better behaved than alternatives and better captures human choice data in two independent datasets of probabilistic learning tasks. The proposed model provides a coherent account of learning in stable or volatile environments and has implications for decision neuroscience research.
Highlights
Our decisions are guided by our ability to associate environmental cues with the outcomes of our chosen actions
Despite the success of statistically founded algorithms for learning in stable environments, in which uncertainty behaves in simple and predictable ways, it is challenging to develop a simple yet efficient algorithm for learning in volatile environments, in which uncertainty dynamically changes over time
According to both psychological and normative models of learning [1,2,3,4], when animals observe pairings between cues and outcomes, they update their belief about the value of the cues in proportion to prediction errors, the difference between the expected and observed outcomes. The degree of this updating depends on a stepsize or learning rate parameter. Some accounts take this as a free parameter, analyses based on statistical inference, such as the Kalman filter [5], instead demonstrate that the learning rate should in principle depend on the learner’s uncertainty
Summary
Our decisions are guided by our ability to associate environmental cues with the outcomes of our chosen actions. A central theoretical and empirical question in behavioral psychology and neuroscience has long been how humans and other animals learn associations between cues and outcomes According to both psychological and normative models of learning [1,2,3,4], when animals observe pairings between cues and outcomes, they update their belief about the value of the cues in proportion to prediction errors, the difference between the expected and observed outcomes. In volatile environments, in which the speed by which true associations change might itself be changing, uncertainty (and learning rates) should fluctuate up and down according to how quickly the environment is changing [6,7] This normative analysis parallels classical psychological theories, such as the Pearce-Hall model [3], which posit that surprising outcomes increase the learning rate while expected ones decrease it. Those models measure surprise by the absolute value of the discrepancy between the actual outcome and the expected value, i.e. the unsigned prediction error [3]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.