TY - JOUR
T1 - PPM-Decay
T2 - A computational model of auditory prediction with memory decay
AU - Harrison, Peter M.C.
AU - Bianco, Roberta
AU - Chait, Maria
AU - Pearce, Marcus T.
PY - 2020/11
Y1 - 2020/11
N2 - Statistical learning and probabilistic prediction are fundamental processes in auditory cognition. A prominent computational model of these processes is Prediction by Partial Matching (PPM), a variable-order Markov model that learns by internalizing n-grams from training sequences. However, PPM has limitations as a cognitive model: in particular, it has a perfect memory that weights all historic observations equally, which is inconsistent with memory capacity constraints and recency effects observed in human cognition. We address these limitations with PPM-Decay, a new variant of PPM that introduces a customizable memory decay kernel. In three studies—one with artificially generated sequences, one with chord sequences from Western music, and one with new behavioral data from an auditory pattern detection experiment—we show how this decay kernel improves the model’s predictive performance for sequences whose underlying statistics change over time, and enables the model to capture effects of memory constraints on auditory pattern detection. The resulting model is available in our new open-source R package, ppm (https://github.com/pmcharrison/ppm).
AB - Statistical learning and probabilistic prediction are fundamental processes in auditory cognition. A prominent computational model of these processes is Prediction by Partial Matching (PPM), a variable-order Markov model that learns by internalizing n-grams from training sequences. However, PPM has limitations as a cognitive model: in particular, it has a perfect memory that weights all historic observations equally, which is inconsistent with memory capacity constraints and recency effects observed in human cognition. We address these limitations with PPM-Decay, a new variant of PPM that introduces a customizable memory decay kernel. In three studies—one with artificially generated sequences, one with chord sequences from Western music, and one with new behavioral data from an auditory pattern detection experiment—we show how this decay kernel improves the model’s predictive performance for sequences whose underlying statistics change over time, and enables the model to capture effects of memory constraints on auditory pattern detection. The resulting model is available in our new open-source R package, ppm (https://github.com/pmcharrison/ppm).
UR - http://www.scopus.com/inward/record.url?scp=85095729060&partnerID=8YFLogxK
U2 - 10.1371/journal.pcbi.1008304
DO - 10.1371/journal.pcbi.1008304
M3 - Journal article
C2 - 33147209
AN - SCOPUS:85095729060
SN - 1553-734X
VL - 16
JO - PLOS Computational Biology
JF - PLOS Computational Biology
IS - 11
M1 - e1008304
ER -