AI Learns to play LUDO | The Joy of Engineering

In a school project developed I an Artificial Intelligence (AI) that was able to play the board game LUDO. The AI used Evolutionary learning with the Genetic Algorithm and a deep neuron network for implementation. In this video I aim to show I implemented it, and how its performance was.

[suwedo]

I've been working with NeuroEvolution algorithms for a few weeks now, and your video was very informative to me, thanks a lot, you really are underrated bro.

[conradhilton]

just found your channel and i gota say keep up the great work and cant wait to see what you do next

[MichaelForestTechnologies]

This video ended up being a long time in the making because of school, and that I choose to experiment with a lot of different video editing / recording techniques. Not all the experimenting turned out that great, so would love to hear what you think of this style of video. :)

[Mhmhind]

Start uploading more videos. Your content is very informative and good

[ЭдуардАндронов-уу]

Very interesting project and a nice video as well: a lot of information in pretty short period of time. Of course a lot of questions appeared right away, but I guess this is sort of an overview, so details are for another video. Like, why adding info about field is so terrifying, wouldn’t it be like “only” 54 or 72 more inputs? What activation function you used? Is “danger” input binary? Is “piece can escape” really needed if we have “can be attacked”? Don’t that cover only a really specific case? Is it worth to be holding on to “under-performers” so much? Why on “figure 2” at 5:00 and later upper neurons have no inputs? And so on.
So, thank you and looking forward to the next video.