Skip to main content

MNEST (Multi-agent Neuro Evolution Simulation Toolkit) is a software framework designed to model and study emergent behavior in complex systems

Project description

MNEST

MNEST (Multi-agent Neuro Evolution Simulation Toolkit) is a software framework designed to model and study emergent behavior in complex systems. The toolkit enables researchers to simulate and analyze multi-agent systems and study their collective behavior, with a particular focus on neuro-evolutionary algorithms, including but not limited to Reinforcement Learning, Q-Learning, Deep-Q, and Genetic Algorithms. The system allows agents to learn from their experiences and evolve over time using Brains that can be configured to be as simple as Q-Tables or as complex as Deep Neural Networks. MNEST provides a visualization tool to help researchers understand the emergent phenomena that arise from the interactions among agents. To improve the research experience, the system also has a Command Line Interface where visualization is disabled to improve speed. The toolkit's modular design and customizable components allow for flexibility and extensibility, making it a versatile tool for exploring the dynamics of complex systems and developing new insights into the behavior of multi-agent systems. The Toolkit is built using OOPs concepts, and hence can be extended to create agents whose limitations lie only in the minds of the creator. Overall, MNEST is an ideal platform for researchers and practitioners seeking to explore the potential of neuro-evolutionary algorithms in the study of emergent behavior in complex systems.

About

Installation.

pip install mnest

Thank you for your patience

This project is under development, I will soon be adding proper documentation and other usage details.

Notes

This was earlier planned to be released as SEAN (A Simulation Environment for Agent-based Neuro-evolution) is a Python package for building and simulating artificial intelligence in multi agent-based models.

Scratch Story.

Everything starts with nothingness. The first thing to be created is the world. The world by its name has no meaning, It still is nothingness. something arives as the first rule for the world is set. Its shape[Rows x Columns].But what good is the shape if it stillhas nothing which is where things start to exist in the world. There are many ways to catagorise the world. here i go with the following structure the world is made of different layers each layer serves a "class" of purpose. E.g. One layer holds the agent , another holds food, etc... The only purpose of the world is to hold the classes of things and five it existance. hence it need not provide the tools for visualisation or calculations. In order to prevent confusions, each cell of a layer can hold only one element. E.g. If ther exist a scanario where say 2 agents can occupy a single position, then each agent is to be given a separate layer.

Now, even before the creation of "thing" or "Agents" which have their own rules and can interact, there must exist 2 more things. Since we(Creators) are playing "GOD", we need a (Gods Perspecitve) method of viusualising the world we created. (Unfotunately in this version of the universe we do not have the ability to reverse time or alter the world once it is set in motion. these are functionalities not too complex to add to the system in other Universes(Versions). even these functionalities are accessible through the gods perspective.) Once we can see what the world looks like, the next thing that needs to exist is a universal clock. One which none of the Universes beings/entities have the ability to control. This clock marches forward one step at a time. the unit of time is irrelavent and can be chosen to please the visualisation and necessities. Thus we have time evolution.

The collection of the clock, the world and the Gods perspective is together called "The Environment"

Project details


Download files

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

Source Distribution

mnest-0.1.0.tar.gz (25.6 kB view details)

Uploaded Source

Built Distribution

mnest-0.1.0-py3-none-any.whl (24.7 kB view details)

Uploaded Python 3

File details

Details for the file mnest-0.1.0.tar.gz.

File metadata

  • Download URL: mnest-0.1.0.tar.gz
  • Upload date:
  • Size: 25.6 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.10.4

File hashes

Hashes for mnest-0.1.0.tar.gz
Algorithm Hash digest
SHA256 1f783ca10cca31cb2b54a1bc25de3728a0899f9835257b6c01f4a2de5a6b30b9
MD5 54c4be2566efd4dd09809bc55dcd68e8
BLAKE2b-256 fa7a18288abdc65a5402c8c98a945558460cfe0eb799ee9289a504c8a27d9e14

See more details on using hashes here.

File details

Details for the file mnest-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: mnest-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 24.7 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.10.4

File hashes

Hashes for mnest-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 50a8d3080cbe3861fa6baac2c5d49ca857f442632193bf02fed00ce7b42b276c
MD5 755dda7faeeb31df4110f31d116541f5
BLAKE2b-256 5e3263caf958e62d788a8f9c2a4d9c0b8bfab3e6b9de8369a9d084724b979504

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page