Perceptual estimation obeys Occam's razor

Abstract

Theoretical models of unsupervised category learning postulate that humans “invent” categories to accommodate new patterns, but tend to group stimuli into a small number of categories. This “Occam's razor” principle is motivated by normative rules of statistical inference. If categories influence perception, then one should find effects of category invention on simple perceptual estimation. In a series of experiments, we tested this prediction by asking participants to estimate the number of colored circles on a computer screen, with the number of circles drawn from a color-specific distribution. When the distributions associated with each color overlapped substantially, participants' estimates were biased toward values intermediate between the two means, indicating that subjects ignored the color of the circles and grouped different-colored stimuli into one perceptual category. These data suggest that humans favor simpler explanations of sensory inputs. In contrast, when the distributions associated with each color overlapped minimally, the bias was reduced (i.e., the estimates for each color were closer to the true means), indicating that sensory evidence for more complex explanations can override the simplicity bias. We present a rational analysis of our task, showing how these qualitative patterns can arise from Bayesian computations.

Discovering latent causes in reinforcement learning

“…For example, the latent cause framework provides insight into the conditions under which old memories are updated and new memories are formed [10]. Similarly, the process of forming and updating memories can be used to understand the unsupervised discovery of visual categories [18]. These applications suggest that structure learning may function as a core computational system that is shared across domains, with the hippocampus and orbitofrontal cortex potentially playing a central role [48,8,10,49,50].…”

“…Most commonly, the prior expresses a simplicity bias (cf. [17,18]) favoring latent structures with a small number of latent causes. Importantly, since in the real world the number of causes is often unknown, the prior must be able to accommodate an unbounded number of them.…”

“…Thus, the number of latent causes can grow as new data are observed, and the parameter a controls the rate of this growth. The Chinese restaurant process has also appeared in many areas of cognitive science, ranging from categorization [42,44] to visual perception [18] and reconstructive memory [10].…”

Occam's Razor in sensorimotor learning

“…In fact, the human sensorimotor system is frequently confronted with such decision-making situations, for instance when perceiving a visual or motion illusion [30,31]. Interestingly, the emergence of many illusions can be explained within the Bayesian framework by using priors that reflect environmental statistics [32][33][34][35], including priors for lightfrom-above illumination in object perception [36,37], priors for low number of categories in object categorization [38] and priors for lower speed in motion perception [30,39]. Deciding for a particular hypothesis can also be thought of as choosing the simpler explanation, which highlights the important role of the prior in defining what is simple or difficult.…”

Reinforcement learning with Marr

“…The advent of principled, statistical methods for studying trial-by-trial learning dynamics [34] has allowed critical examination of the early phases of learning, when animals learn the ‘rules of the game’ (or the state representation). Behavioral findings point to a process akin to Bayesian inference in which animals attempt to use observed information to infer the unobservable (latent) causal structure of the task, which is then used to craft task states that accurately describe the task dynamics [35–38]. This ‘representation learning’ process has been linked to memory processes [39,40] as well as neural selective attention mechanisms [41].…”