9.10: Genetic Algorithm: Continuous Evolutionary System – The Nature of Code

[ad_1]
In this video, I apply the Genetic Algorithm to an “Ecosystem Simulation”, a system in which models biological life more closely, where elements live and die continuously evolving over time.

Support this channel on Patreon:

Send me your questions and coding challenges!:

Contact:

Source Code for the Video Lessons:

p5.js:
Processing:

For More Genetic Algorithm videos:

For More Nature of Code videos:

Help us caption & translate this video!


Posted

in

by

Tags:

Comments

30 responses to “9.10: Genetic Algorithm: Continuous Evolutionary System – The Nature of Code”

  1. Bilal Haider Avatar

    the whole time, i was watching those bloops :p and i did not listened to what you were blabbering about 😀
    awesome work..

  2. Shadid Haque Avatar

    are you the God of smart rockets

  3. Aharon Jacobson Avatar

    Are you going to make videos about neural networks? You said you would in a video didn't you?

  4. mahdi rafatjah Avatar

    +shiffman Man you know to have a cross over we need at least 2 genome at DNA !

  5. Elias070 Avatar

    Hook up a neural network

  6. SEFD Game Dev Avatar

    personal opinion here, but I think ultimately neural net-less continuous genetic algorithms or non-user input GAs from my experiences are… bleh. There just is something about setting a goal in the form of fitness which dramatically increases performance. Having random mutations with hard coded parameters are hard to get exciting results from, or so ive found from my projects, I'd be interested to see results from someone else. An extremely intricately crafted fitness scoring convention is a much better demonstration of Evolution I believe, & so instead I would suggest to anyone looking to create any continuous evolutionary systems to instead apply a Dynamic GA instead.

    So i.e. every attempt at hunting is an element of n1, every attempt at mating is an element of n2, so on & so fourth, & so when it comes time to reproduce, you would do parent[0].n + parent[1].n = N,parent[0].m + parent[1].m = M, …. then use N & M as your mating pool to determine child DNA.

    A little more depth involved, but I think you'll find for more exciting results!

  7. julien jamet Avatar

    Gros merci aux traducteurs !!!

  8. John David Deatherage Avatar

    I've watched your GA series of videos. I'm teaching myself Python and I decided to try and code a GA to evolve to match a phrase. I've written two CrossOver functions called weave and top2bottom. Most videos I've watched call for a fitness calculation and then delete a percentage of the population (half). And then like human reproduction, we take two parents to produce an offspring. Here's my question. Why limit a GA? If you have a population of 1,000, why not match each population member with every other member (think nested loops)? A population of 1000 would produce a 1,000,000. Run CrossOver, then the mutation function. Rank the population by fitness. Then run the grim reaper function to select only the top 1000 (1%) by fitness and repeat the process until you've reached the solution? I'd like your opinion of my idea. Thank you for taking the time to read and answer my question. (I'm open to suggestions from anyone…..)

  9. özge özçelik Avatar

    STATS latin square design gerechte designs

  10. William Antônio Siqueira Avatar

    Great videos, as always! You help me to keep alive the joy of programming.

    I can't find the code for this lesson! Would you please add all the genetic algorithm code to a specific repo or a directory? You are the best teacher ever.

  11. Ricard Vilanova Avatar

    How do the "bloops" move around the screen looking for food?
    Do they have their own genetic algorithm looking for the best path to the food or it's just some random direction?

  12. pow3rstrik3 Avatar

    So technically the bloops could go extinct…..

  13. Trenton Streator Avatar

    Please release the source code for the continuous Evolutionary system. I would like to experiment and create a working ecosystem from several creatures, and I'm having a hell of a time re-writing the code in Visual C#. I would appreciate the source code to help me translate a little bit. The algorithms involved and code structure would help a lot.

  14. Gurpreet Singh Avatar

    Great videos and a great style….

  15. jimmy burton Avatar

    thank you senpai. what a wonderful thing this is – your style and technique of teaching resonates with me so well. It's clear you have a gift, thank you for sharing it!

  16. Jowanza Joseph Avatar

    These videos are the best.

  17. MeLlamoJeffe Avatar

    Will you go into something as complex as neural networks?

  18. Kyle Amoroso Avatar

    Oh cool, so this is like asexual reproduction. I think a cool way to model the reproduction would be to have a higher chance of reproducing if it reaches a certain amount of food, then halve it's lifespan. Kind of like a cell gathering resources to reproduce then splitting in two.

  19. CAMILOL Avatar

    Hi dan, can you make a coding challenge? coding a spectre sound in processing with minim. thank you.

  20. Irohin Avatar

    Really interesting! Loving this channel :))

  21. Piotr Kraso Avatar

    how can I acces the source code for the genetic algorithms?

  22. S Avatar

    Really helpful channel

  23. Richard Brown Avatar

    Here is my contribution to the bloopiverse. I have tried to extend the scope of the DNA to control appearance and movement of the 'cells' and also introduced sex & strains (skipped food, though).
    https://rik-brown.github.io/Cellendipity_NOC/ (it was my final submission for the recent Kadenze 'Nature of Code' course, which I see is running again in September – highly recommended!) https://www.kadenze.com/courses/the-nature-of-code/info Code here: https://github.com/rik-brown/Cellendipity_NOC

  24. Jack Attacker Avatar

    I new episode you make my day!!! 😀

  25. Robert Coyle Avatar

    can you go on with an example of a two ( multiple) gene system… Say a gene for size/speed and a gene for mouth parts/ size/location. How would you combine them as an overall fitness?

  26. Sreekanth Narayanaswamy Avatar

    Why does it say August 11 ? I thought this ws uploaded way back . I am having a de-ja-vu moment here. !

  27. TheWGBbroz Avatar

    Noo I missed the live stream D:

  28. Matthias Krijgsman Avatar

    I really love these video's, they always help me out, Thanks!

Leave a Reply

Your email address will not be published. Required fields are marked *