Abstract
The capability to establish accurate predictions is an integral part of learning. Whether predictions about different dimensions of a stimulus interact with each other, and whether such an interaction affects learning, has remained elusive. We conducted a statistical learning study with EEG (electroencephalography), where a stream of consecutive sound triplets was presented with deviants that were either: (a) statistical, depending on the triplet ending probability, (b) physical, due to a change in sound location or (c) double deviants, i.e. a combination of the two. We manipulated the predictability of stimulus-onset by using random stimulus-onset asynchronies. Temporal unpredictability due to random onsets reduced the neurophysiological responses to statistical and location deviants, as indexed by the statistical mismatch negativity (sMMN) and the location MMN. Our results demonstrate that the predictability of one stimulus attribute influences the processing of prediction error signals of other stimulus attributes, and thus also learning of those attributes.
Highlights
Learning is a fundamental property of nervous systems, and recent accounts from cognitive and computational neuroscience have established the underlying role of prediction for learning [1,2,3,4,5]
We used the mismatch negativity (MMN) to investigate how predictions based on statistical learning are affected when predictability of stimulus-onset is manipulated by using random stimulus-onset asynchronies
The observation that no statistical mismatch negativity (sMMN) was elicited under non-isochronous stimulation reveals the different nature of the neural traces and the functional operations engaged during the elicitation of the sMMN compared to the “classical” MMN observed in traditional oddball paradigms [19, 20, 22, 23, 40]
Summary
Learning is a fundamental property of nervous systems, and recent accounts from cognitive and computational neuroscience have established the underlying role of prediction for learning [1,2,3,4,5]. Within the framework of predictive coding (PC), learning is the generation of a predictive model of the world which is continually optimized by reducing prediction errors, i.e. any mismatches between incoming sensory signals and the predicted encoded input [6, 7]. In real life we tend to learn by forming predictions over several dimensions (e.g., location, intensity, timing). Our predictive model encompasses a multitude of features, and it remains unknown if predictions about one feature affect predictions about another feature, and whether this plausible interaction affects model precision and the actual learning. A neurophysiological marker of model precision is the mismatch negativity (MMN), which is an electrical brain response to a deviant auditory stimulus among standards [8]. We used the MMN to investigate how predictions based on statistical learning are affected when predictability of stimulus-onset is manipulated by using random stimulus-onset asynchronies
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: 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.
