We are interested in the brain as an exemplar of "biological intelligence". That is, in contrast to current autonomous machines, the brain deals efficiently, possibly optimally, with noisy information and is not only able to make quick inferences based on ambiguous tasks, but also encode the degree of associated uncertainty.
We are seeking to explore this using simple perceptual tasks, such as vibrotactile discrimination, but embedding increasing levels of ambiguity in the stimuli and task requirements. Our goal is to understand how the brain deals with the perceptual uncertainty and where its solution lies in the decision space of efficiency versus accuracy. Using functional neuroimaging and computational modelling, we then seek the neural instantiations of these processes.
Although there are many other aspects to decision making, we are using decisions made on these
perceptually ambiguous tasks in order to understand the nature and source of ambivalence, a key and disabling cognitive symptom of both depression and schizophrenia.