From statistical inference to a differential learning rule for stochastic neural networks. Issue 6 (6th December 2018)

• Record Type: Journal Article
• Authors: Saglietti, Luca; Gerace, Federica; Ingrosso, Alessandro; Baldassi, Carlo; Zecchina, Riccardo
• Is Part Of: Interface focus. Volume 8:Issue 6(2018)
• Publication Date: 2018-12-06
• Subjects: associative memory -- attractor networks -- learning; Physical sciences -- Periodicals; Life sciences -- Periodicals; 500
• DOI: 10.1098/rsfs.2018.0033
• Languages: English
• ISSNs: 2042-8898

Abstract

Abstract : Stochastic neural networks are a prototypical computational device able to build a probabilistic representation of an ensemble of external stimuli. Building on the relationship between inference and learning, we derive a synaptic plasticity rule that relies only on delayed activity correlations, and that shows a number of remarkable features. Our delayed-correlations matching (DCM) rule satisfies some basic requirements for biological feasibility: finite and noisy afferent signals, Dale's principle and asymmetry of synaptic connections, locality of the weight update computations. Nevertheless, the DCM rule is capable of storing a large, extensive number of patterns as attractors in a stochastic recurrent neural network, under general scenarios without requiring any modification: it can deal with correlated patterns, a broad range of architectures (with or without hidden neuronal states), one-shot learning with the palimpsest property, all the while avoiding the proliferation of spurious attractors. When hidden units are present, our learning rule can be employed to construct Boltzmann machine-like generative models, exploiting the addition of hidden neurons in feature extraction and classification tasks.