Growing dendrites enhance a neuron's computational power and memory capacity

TL;DR

This paper introduces a novel neuro-inspired algorithm that leverages dendritic growth and synaptogenesis, enhancing a neuron's memory, avoiding forgetting, and enabling it to unmix complex data distributions, with potential for improved generalization.

Contribution

The paper presents a new stochastic, Hebbian-based algorithm combining dendritogenesis and supervised synaptogenesis, demonstrating enhanced computational and memory capabilities in neurons.

Findings

Neurons with this algorithm have increased memory capacity.

The algorithm prevents catastrophic forgetting.

It can unmix mixture distributions effectively.

Abstract

Neocortical pyramidal neurons have many dendrites, and such dendrites are capable of, in isolation of one-another, generating a neuronal spike. It is also now understood that there is a large amount of dendritic growth during the first years of a humans life, arguably a period of prodigious learning. These observations inspire the construction of a local, stochastic algorithm based on an earlier stochastic, Hebbian developmental theory. Here we investigate the neuro-computational advantages and limits on this novel algorithm that combines dendritogenesis with supervised adaptive synaptogenesis. Neurons created with this algorithm have enhanced memory capacity, can avoid catastrophic interference (forgetting), and have the ability to unmix mixture distributions. In particular, individual dendrites develop within each class, in an unsupervised manner, to become feature-clusters that correspond to the mixing elements of class-conditional mixture distribution. Although discriminative problems are used to understand the capabilities of the stochastic algorithm and the neuronal connectivity it produces, the algorithm is in the generative class, it thus seems ideal for decisions that require generalization, i.e., extrapolation beyond previous learning.