Skip to content

Tuning the Parameters of Heuristic Optimizers (Meta-Optimization / Hyper-Parameter Optimization)

License

Notifications You must be signed in to change notification settings

Hvass-Labs/MetaOps

Repository files navigation

MetaOps

Original repository on GitHub

Original author is Magnus Erik Hvass Pedersen

Introduction

This is a small collection of research papers on automatic tuning of the parameters of a heuristic optimizer such as Genetic Algorithm, Particle Swarm Optimization, and Differential Evolution. The parameter tuning is done by another overlaying optimizer and this is often called Meta-Optimization or Hyper-Parameter Optimization.

Papers

  1. Bayesian Meta-Optimization (Notebook) (Google Colab)

  2. Multi-Objective Meta-Optimization (Notebook) (Google Colab)

  3. Meta-Optimization Using Many Problems (Notebook) (Google Colab)

Videos

There is a YouTube video for each research paper.

Downloading

It is recommended that you download the whole repository from GitHub, instead of just downloading the individual Python Notebooks.

Git

The easiest way to download and install this is by using git from the command-line:

git clone https://github.com/Hvass-Labs/MetaOps.git

This creates the directory MetaOps and downloads all the files to it.

This also makes it easy to update the files, simply by executing this command inside that directory:

git pull

Zip-File

You can also download the contents of the GitHub repository as a Zip-file and extract it manually.

How To Run

If you want to edit and run the Notebooks on your own computer, then it is suggested that you use the Anaconda distribution with Python 3.6 (or later) because it has most of the required packages already installed. Then you type the following commands in a terminal window:

cd MetaOps
conda create --name metaops python=3.6
source activate metaops
pip install -r requirements.txt

Now you can run the Notebooks by typing this command:

jupyter notebook

Run in Google Colab

If you do not want to install anything on your own computer, then the Notebooks can be viewed, edited and run entirely on the internet by using Google Colab. You can click the "Google Colab"-link next to the research papers listed above. You can view the Notebook on Colab but in order to run it you need to login using your Google account. Then you need to execute the following commands at the top of the Notebook, which clones MetaOps to your work-directory on Colab and installs the required packages.

import os
work_dir = "/content/MetaOps/"
if os.getcwd() != work_dir:
    !git clone https://github.com/Hvass-Labs/MetaOps.git
os.chdir(work_dir)
!pip install -r requirements.txt

License (MIT)

These Python Notebooks and source-code are published under the MIT License which allows very broad use for both academic and commercial purposes.

You are very welcome to modify and use the source-code in your own project. Please keep a link to the original repository.

Releases

No releases published

Packages

No packages published