EPOS - Economic Planning and Optimized Selections

Contents:

• Initial Setup and Running Experiments
  • Command-Line Setup
  • IDE Setup
• Configuration Parameters
• Input
  • Plans
  • Incentive Signals
• Global Cost Function
  • Variance
  • Cross-Correlation
  • Residual Sum of Squares
  • Root Mean Square Error
  • Scaling Techniques
• Local Cost Function
  • Index-based Local Cost Function
  • Discomfort-based Local Cost Function
  • Preference-based Local Cost Function
• Tree Topology
• Termination Criteria
• Multi-Objective Optimization
• Logging & Output
  • Global Complex Cost Logger
  • Global Cost Logger
  • Global Response Logger
  • Indexed Histogram Logger
  • Local Cost Logger
  • Selected Plans
  • Termination Logger
  • Unfairness Logger
  • Reorganization Logger
  • Weights, Alpha, Beta Logger
• Advanced Topics
  • Tree Reorganization
  • Optimality
  • How to run experiments with EPOS and COHDA
  • How to run live
  • How to write your own agent

Installation

The project can be installed in 2 possible ways:

• A command line executable, that can be found under the project github releases. Once the release-x.x.x.zip is downloaded and extracted the user can navigate in the extracted folder and follow the guideline at Command-Line Setup. For this setup a Java 8 or higher should be installed.
• IDE imported source code, that can be found under the tutorial branch of the project github repository. Detailed instructions for getting and importing the source code to an IDE can be found at IDE Setup. For this setup jdk-8uxxx or higher should be installed along with Maven 3.3 or higher and the latest version of git.

Open Datasets

Energy: http://epos-net.org/shared/datasets/EPOS-ENERGY-SUBSET.zip

Electric vehicles: http://epos-net.org/shared/datasets/EPOS-ELECTRIC-VEHICLES.zip

Bike sharing: http://epos-net.org/shared/datasets/EPOS-BICYCLES.zip

Software Artifact

An earlier software artifact with a Graphical User Interface (GUI) can be acquired via EPOS Exemplar. The website contains an executable and instructions about the latest artifact version of EPOS. It is advised to use the website software for demonstration purposes. Detailed usage guidelines and explanations regarding the system architectures can also be found in the website.

Contribute

• Issue Tracker
• Source Code

Support

For any question please contact us: mail@epos-net.org

License

The project is licensed under the GPL-2.0 license.

Publications

Evangelos Pournaras, Peter Pilgerstorfer and Thomas Asikis, Decentralized Collective Learning for Self-managed Sharing Economies, ACM Transactions of Autonomous and Adaptive Systems, 2018, ACM, url

Indices and tables

• Index
• Module Index
• Search Page