Non-linear Neurons with Human-like Apical Dendrite Activations
Authors:
- Mariana-Iuliana GeorgescuMGMariana-Iuliana Georgescu
- Radu Tudor IonescuRIRadu Tudor Ionescu
- Nicolae-Catalin RisteaNRNicolae-Catalin Ristea
- Nicu SebeNSNicu Sebe
Preprint
Abstract
In order to classify linearly non-separable data, neurons are typically organized into multi-layer neural networks that are equipped with at least one hidden layer. Inspired by some recent discoveries in neuroscience, we propose a new neuron model along with a novel activation function enabling learning of non-linear decision boundaries using a single neuron. We show that a standard neuron followed by the novel apical dendrite activation (ADA) can learn the XOR logical function with 100% accuracy. Furthermore, we conduct experiments on three benchmark data sets from computer vision and natural language processing, i.e. Fashion-MNIST, UTKFace and MOROCO, showing that the ADA and the leaky ADA functions provide superior results to Rectified Liner Units (ReLU) and leaky ReLU, for various neural network architectures, e.g. 1-hidden layer or 2-hidden layers multi-layer perceptrons (MLPs) and convolutional neural networks (CNNs) such as LeNet, VGG, ResNet and Character-level CNN. We also obtain further improvements when we change the standard model of the neuron with our pyramidal neuron with apical dendrite activations (PyNADA).Keywords
All Related Versions
This publication has 9 references indexed in Scilit:
- Dendritic action potentials and computation in human layer 2/3 cortical neuronsScience, 2020
- Deep appearance and motion learning for egocentric activity recognitionNeurocomputing, 2018
- Towards deep learning with segregated dendriteseLife, 2017
- Detecting anomalous events in videos by learning deep representations of appearance and motionComputer Vision and Image Understanding, 2017
- Why Neurons Have Thousands of Synapses, a Theory of Sequence Memory in NeocortexFrontiers in Neural Circuits, 2016
- The Lure of the UnknownNeuron, 2006
- Approximate Statistical Tests for Comparing Supervised Classification Learning AlgorithmsNeural Computation, 1998
- Learning representations by back-propagating errorsNature, 1986
- The perceptron: A probabilistic model for information storage and organization in the brain.Psychological Review, 1958