2021/06/30 by Thomas R. Shultz, Shultz, Thomas R, Ardavan Salehi Nobandegani +1
Computer Science · Psychology · #Bayesian Modeling and Causal Inference #Child and Animal Learning Development #FOS: Biological sciences #Neural Networks and Applications #Neurons and Cognition (q-bio.NC)
paper · pdf · doi:10.48550/arxiv.2106.16059
openalex publication_date 2021/06/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Recent experiments reveal that 6- to 12-month-old infants can learn\nprobabilities and reason with them. In this work, we present a novel\ncomputational system called Neural Probability Learner and Sampler (NPLS) that\nlearns and reasons with probabilities, providing a computationally sufficient\nmechanism to explain infant probabilistic learning and inference. In 24\ncomputer simulations, NPLS simulations show how probability distributions can\nemerge naturally from neural-network learning of event sequences, providing a\nnovel explanation of infant probabilistic learning and reasoning. Three\nmathematical proofs show how and why NPLS simulates the infant results so\naccurately. The results are situated in relation to seven other active research\nlines. This work provides an effective way to integrate Bayesian and\nneural-network approaches to cognition.\n