MONCAE: Multi-Objective Neuroevolution of Convolutional Autoencoders
arXiv:2106.11914
Abstract
In this paper, we present a novel neuroevolutionary method to identify the architecture and hyperparameters of convolutional autoencoders. Remarkably, we used a hypervolume indicator in the context of neural architecture search for autoencoders, for the first time to our current knowledge. Results show that images were compressed by a factor of more than 10, while still retaining enough information to achieve image classification for the majority of the tasks. Thus, this new approach can be used to speed up the AutoML pipeline for image compression.
arXiv commentPublished as a Poster paper in ICLR 2021 Neural Architecture Search workshop
References in corpus (13)
- Deep Learning in Neural Networks: An Overview
- Distilling the Knowledge in a Neural Network
- Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
- Neural Architecture Search with Reinforcement Learning
- AutoML: A Survey of the State-of-the-Art
- The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks
- A Widely Applicable Bayesian Information Criterion
- Medical image denoising using convolutional denoising autoencoders
- Generating Images with Perceptual Similarity Metrics based on Deep Networks
- Large image datasets: A pyrrhic win for computer vision?
- NASNet: A Neuron Attention Stage-by-Stage Net for Single Image Deraining
- Does Unsupervised Architecture Representation Learning Help Neural Architecture Search?
- EvoAAA: An evolutionary methodology for automated \neural autoencoder architecture search