Skip to main content
Join the official Python Developers Survey 2018 and win valuable prizes: Start the survey!

Swarm intelligence hyperparameter optimization.

Project description

particles

SwarmOpt

SwarmOpt is a swarm intelligence optimizer for hyperparameter tuning. The project is currently in R&D, where I'm implementing swarm intelligence algorithms to find the best variation for hyperparemeter tuning. I'm particularly interested in comparing velocity clamp variations, and implementing test functions that most closely resemble the kinds of complex search environments encountered in hyperparameter tuning.

Swarm intelligence leverages global population-based search solutions to balance exploration and exploitation with respect to specified cost functions. There are many exciting nooks and crannies to explore in the SI toplogy, yet I've chosen to kick things of with some Particle Swarm Optimization (PSO) algorithms, as they are easy to understand and fiddle with. The PSO lineage was sparked by Eberhart and Kennedy in their original paper on PSOs in 1995, and the intervening years have seen many variations spring from their central idea.

Algorithms

Benchmark Functions

Single objective test functions:

  • Sphere Function
  • Rosenbrock's Function
  • Ackley's Function
  • Griewank's Function
  • Rastrigin's Function
  • Weierstrass Function

On Deck

  • Cooperative Approach to PSO (CPSO)(multiple collaborating swarms)
  • Proactive Particles in Swarm Optimization (PPSO) (self-tuning swarms)
  • Inertia weight variations
  • Mutation operator variations
  • Velocity clamping variations
  • Multiobjective variations
  • Dynamic environment

Applications

  • Neural network number of layers and structure optimization
  • Grid scheduling (load balancing)
  • Routing in communication networks

Citation

Siobhán K Cronin, SwarmOpt (2018), GitHub repository, https://github.com/SioKCronin/SwarmOpt

Project details


Release history Release notifications

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Filename, size & hash SHA256 hash help File type Python version Upload date
swarmopt-0.1.0-py3-none-any.whl (3.8 kB) Copy SHA256 hash SHA256 Wheel py3 Jun 15, 2018
swarmopt-0.1.0.tar.gz (3.3 kB) Copy SHA256 hash SHA256 Source None Jun 15, 2018

Supported by

Elastic Elastic Search Pingdom Pingdom Monitoring Google Google BigQuery Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN SignalFx SignalFx Supporter DigiCert DigiCert EV certificate StatusPage StatusPage Status page