Page Tag-Cloud
Software Tag-Cloud
Building from Source
Open Source Definition
All Software
  Popular Tags
C Plus Plus
Source Code
Notable Members
Our Company
Copyright Information
Software EULA
Pre-Release EULA
Privacy Policy
Make Contact
Downloads   0
User Rating   (Rate)
Last Updated   n/a
License   Apache License
Recommended Release
No recommended release for this software.

View all Releases

This software is open source. You can obtain the latest source code from the GitHub repository or browse the releases for the source code associated with a specific release. If you make any changes which you feel improves this application, please let us know via our Contact Page.

This has been replaced with the C# version here: https://github.com/NTDLS/AIVolution

AIEvolution is an evolution simulator based on evolutionary specialization using our neural network implementation known as DetermiNet.

Evolution in plain terms: The theory behind evolution is based around the scientific evidence that creatures change randomly from time to time (due to genetics, radiation, you name it). Some of these changes are detrimental to the degree that the creature's blood-line will off within a relatively short period of time (or is unable to reproduce), however... in rare occasions these mutations can be beneficial to the degree that the new brand of creature will rise to the top of his particular species - or at the very least: her new blood-line will "out survive" that of her nearest relative species.

AIEvolution attempts to simulate these evolutionary happenings through the use of random chance, rule based survival and the forward generational propagation of all traits (good and bad) from a sample of supposed superior virtual-creatures. This operation is repeated in cycles (known as generations) for a set number of iterations and then the traits (and resulting behaviors) of the ending generation are compared in detail to those of the beginning generation.

Main window of AIEvolution:


Rules of the virtual universe:
  1. Directional movement in free-space must be random.
  2. Blockades are in random positions and must be honored.
  3. Resource pools are in random positions and supply creatures within close proximity with resource points.
  4. Each time an interaction with resource or another creature is required, the creature will face decisions and exhibit one or more built-in behaviors which are driven by individualized artificial neural networks. Behaviors include:
    • Loner: When resource is encountered, the creature will stop and collect.
    • Helpful: When resource is encountered, the creature will stop, collect and alert other creature in the proximity of the find.
    • Liar: When resource is encountered, the creature will stop, collect and inform other creature in the proximity that there is nothing in the area - suggesting the proximity creature goes in a different direction.
    • Accepting: A creature that is alerted to nearby resource will go to that resource and collect
    • Arrogance: A creature that is alerted to nearby resource will avoid that resource.
    • Untrusting: A creature which is informed that there is nothing in the area by a resource collecting creature will go to that resource and collect.
  5. After all configured resource is exhausted, the generation will end.
  6. The top 50% of creatures with the most collected resource points will have their traits carried forward to the next generation.
    • Traits are carried forward by randomly pairing the top performers via the merger of their neural networks.
    • The pairing will continue until the original number of creatures has been generated.
  7. The next generation will begin and the cycle will be repeated a configured number of times.

 Academia    AI    Experimental    Neural Network    Showcase    Simulation  

No comments currently exists for this software. Why don't you add one?
First Previous Next Last 

Copyright © 2023 NetworkDLS.
All rights reserved.
Privacy Policy | Our Company | Contact