Skip to content

An agent-based approach to machine-learning demonstrating the use of agentMET4FOF for such applications.

Notifications You must be signed in to change notification settings

Met4FoF/agentMET4FOF_ml

Repository files navigation

Met4FoF use case machine learning

This is supported by European Metrology Programme for Innovation and Research (EMPIR) under the project Metrology for the Factory of the Future (Met4FoF), project number 17IND12.

Purpose

This is an implementation of an agent-based approach to machine learning utilizing the external Python library scikit-multiflow.

Getting started

First you need to create a virtual environment, install all dependencies and finally execute the examples. To install the dependencies you first have to install numpy and then the specified requirements from the requirements.txt.

Create a virtual environment on Windows

In your Windows command prompt execute the following to set up a virtual environment in a folder of your choice.

> python -m venv agentMET4FOF_ml_env
> agentMET4FOF_ml_env\Scripts\activate.bat
(agentMET4FOF_ml_env) > pip install --upgrade pip numpy

And then install the required dependencies after navigating to the root of your clone.

(agentMET4FOF_ml_env) > pip install -r requirements.txt

Create a virtual environment on Mac and Linux

In your terminal execute the following to set up a virtual environment in a folder of your choice.

$ python3 -m venv agentMET4FOF_ml_env
$ source agentMET4FOF_ml_env/bin/activate
(agentMET4FOF_ml_env) $ pip install --upgrade pip numpy

And then install the required dependencies after navigating to the root of your clone.

(agentMET4FOF_ml_env) $ pip install -r requirements.txt

In case you are using PyCharm, you will already find proper run configurations at the appropriate place in the IDE. If not you can proceed executing the script files manually.

If you have any questions please get in touch with the author.

Scripts

The interesting parts you find in the subfolder agentMET4FOF_ml.

Orphaned processes

In the event of agents not terminating cleanly, you can end all Python processes running on your system (caution: the following commands affect all running Python processes, not just those that emerged from the agents).

Killing all Python processes in Windows

In your Windows command prompt execute the following to terminate all python processes.

> taskkill /f /im python.exe /t
>

Killing all Python processes on Mac and Linux

In your terminal execute the following to terminate all python processes.

$ pkill python
$

References

For details about the agents refer to the upstream repository agentMET4FOF

Screenshot of web visualization

Web Screenshot

About

An agent-based approach to machine-learning demonstrating the use of agentMET4FOF for such applications.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages