there is tetris on js, there is a neural network on js that can learn to do or xor and operations.

Is it possible to combine them, or do we need something else?

Closed due to the fact that the question is not clear to the participants by pavel , user207618, aleksandr barakin , katso , post_zeew Oct 22 '16 at 16:00 .

Try to write more detailed questions. To get an answer, explain what exactly you see the problem, how to reproduce it, what you want to get as a result, etc. Give an example that clearly demonstrates the problem. If the question can be reformulated according to the rules set out in the certificate , edit it .

    1 answer 1

    Theoretically yes:

    We define a matrix with an active figure, and then a window (i.e., the dimensions of this matrix) move along the top of the already established figures.

    Thus summ += a[i][j] !== b[i][j] , the place where summ greater is the place to put the figure. Something like an inverted image recognition.

    Training is the flow to the input of the figure, and the output is the inverted value. Thus, the network will be trained to look for an inverted image. The error can be considered according to the formula I gave, it will look for the distance (the number of differences) between the figures.

    This is a general, rather superficial description of an approach that can be based on Boolean operations, which can be used in your case.

    In general, I am not an expert in neural networks)