AI for a card game?

Question

Game development

Given:
there is an implementation of a card game via the Internet. Live people play, connect in pairs and go ahead. Platform - Windows (.NET). Everything is fine, everything works zashib.
Want:
Write a robot to play the same game. But not stupid, but to train protein players on the game. Of course, the server already has full info on each game: on the cards of each player, on the moves and on the result - won / lost. That is, the material is.
But in ML I'm full of null or nil . I read it, I realized that I kind of need " Behavioural Cloning ". Or " Reinforcement Learning ". Basically, I don't even know where to start. What to read, what to choose?
The server is running on Azure. Surely they have something ML. Has anyone used it?

Answer 1

If indeed "But in ML I am completely null or nil." then the answer to the question "where to start" is self-evident. Start from the beginning, by learning ML. From the basics.
What to read, what to choose? - this question was asked 284 times even on the Toaster. I understand that I'm too lazy to search, read, analyze myself. I wish someone would do it for us. Well, okay - here:
https://tproger.ru/curriculum/data-scientist-curri...
www.cyberforum.ru/ai/thread2010374.html#post10584290
https://proglib.io/p/data- science-faq/
ai-news.ru/2018/10/podborka_resursov_po_mashinnomu...
there is a LOT of information about how to start, and in what order to study, and which books, sites, videos to watch. Choose depending on the level of your training, tastes and mood.
Surely they have something ML. - they have it, who has it? And what exactly "is"? Yes, there is an "Azure Machine Learning service" on Azure . But what do you want to find there? And how are you going to use it, without even understanding what you need?
Perhaps someone has already used it - What, "Azure"? Yes, they did. I think tens of thousands of people. So what? How will this information help you solve your problem?
In general, it is not clear what exactly you expect to hear in response to your question.