A Study of Genetic Algorithms for Hyperparameter Optimization of Neural Networks in Machine Translation

09/15/2020
by   Keshav Ganapathy, et al.
0

With neural networks having demonstrated their versatility and benefits, the need for their optimal performance is as prevalent as ever. A defining characteristic, hyperparameters, can greatly affect its performance. Thus engineers go through a process, tuning, to identify and implement optimal hyperparameters. That being said, excess amounts of manual effort are required for tuning network architectures, training configurations, and preprocessing settings such as Byte Pair Encoding (BPE). In this study, we propose an automatic tuning method modeled after Darwin's Survival of the Fittest Theory via a Genetic Algorithm (GA). Research results show that the proposed method, a GA, outperforms a random selection of hyperparameters.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
01/12/2019

Recombination of Artificial Neural Networks

We propose a genetic algorithm (GA) for hyperparameter optimization of a...
research
02/24/2021

A Genetic Algorithm with Tree-structured Mutation for Hyperparameter Optimisation of Graph Neural Networks

In recent years, graph neural networks (GNNs) have gained increasing att...
research
01/07/2021

An automated machine learning-genetic algorithm (AutoML-GA) approach for efficient simulation-driven engine design optimization

In recent years, the use of machine learning techniques as surrogate mod...
research
05/31/2019

Prediction and optimization of mechanical properties of composites using convolutional neural networks

In this paper, we develop a convolutional neural network model to predic...
research
06/23/2020

Efficient Hyperparameter Optimization in Deep Learning Using a Variable Length Genetic Algorithm

Convolutional Neural Networks (CNN) have gained great success in many ar...
research
05/15/2021

FOGA: Flag Optimization with Genetic Algorithm

Recently, program autotuning has become very popular especially in embed...
research
04/28/2022

Automatic Machine Learning for Multi-Receiver CNN Technology Classifiers

Convolutional Neural Networks (CNNs) are one of the most studied family ...

Please sign up or login with your details

Forgot password? Click here to reset