Bio-inspired Machine Learning: programmed death and replication

TL;DR

This paper explores bio-inspired algorithms mimicking biological replication and programmed death to enhance neural network performance and efficiency, demonstrating their effectiveness through experiments on the MNIST dataset.

Contribution

Introduces novel bio-inspired algorithms for neuron addition and removal, improving neural network compression and performance, and combining these methods for better learning efficiency.

Findings

Programmed death algorithm effectively compresses neural networks.

Replication algorithm enhances neural network performance.

Combined algorithms improve learning efficiency on MNIST.

Abstract

We analyze algorithmic and computational aspects of biological phenomena, such as replication and programmed death, in the context of machine learning. We use two different measures of neuron efficiency to develop machine learning algorithms for adding neurons to the system (i.e. replication algorithm) and removing neurons from the system (i.e. programmed death algorithm). We argue that the programmed death algorithm can be used for compression of neural networks and the replication algorithm can be used for improving performance of the already trained neural networks. We also show that a combined algorithm of programmed death and replication can improve the learning efficiency of arbitrary machine learning systems. The computational advantages of the bio-inspired algorithms are demonstrated by training feedforward neural networks on the MNIST dataset of handwritten images.