Validation of Bayesian strategy in probabilistic inference by evaluating the ability to generalise knowledge
Speaker: Dr Sophie Lin
30th March 2023
Abstract: Numerous studies have found that the Bayesian framework, which formulates the optimal integration of the knowledge of the world (i.e. prior) and current sensory evidence (i.e. likelihood), captures human behaviours sufficiently well. However, there are debates regarding whether humans use precise but cognitively demanding Bayesian computations for behaviours. Across two studies, we trained participants to estimate hidden locations of a target drawn from priors with different levels of uncertainty. In each trial, scattered dots provided noisy likelihood information about the target location. Participants showed that they learned the priors and combined prior and likelihood information to infer target locations in a Bayes-fashion. We then introduced a transfer condition presenting a trained prior and a likelihood that have never been put together during training. How well participants integrate this novel likelihood with their learned prior is an indicator of whether participants perform Bayesian computations. In one study, participants experienced the newly introduced likelihood, which was paired with a different prior, during training. Participants changed likelihood weighting following expected directions although the degrees of change were significantly lower than Bayes-optimal predictions. In another group, the novel likelihoods were never used during training. We found people integrated a new likelihood within (interpolation) better than the one outside (extrapolation) the range of their previous learning experience and were quantitatively Bayes-suboptimal. We replicated the findings of both studies in a validation dataset. Our results showed that Bayesian behaviours may not always be achieved by a full Bayesian computation. Future studies can apply our approach in different tasks to enhance the understanding of decision-making mechanisms.
Bio: Sophie is a postdoctoral researcher in Marta Garrido's Cognitive Neuroscience and Computational Neuropsychiatry Lab at the University of Melbourne, developing the first wearable magnetoencephalography (MEG) in Australia and investigating decision-making using the Bayesian framework. Previously, she worked as a research fellow in 2017-2019, under the supervision of Chris Miall (University of Birmingham) and Gareth Barnes (University College London), applying wearable MEG to investigate learning signals in the human cerebellum. She did her PhD in computational neuroscience at Imperial College London. Supervised by Aldo Faisal, she used Bayesian decision theory and psychophysics to relate increased sensorimotor uncertainty to a higher risk of fall incidents in older people. Not by coincidence, she is also a neurologist trained in Taiwan.