How to properly train a neural network model?

Question

Python

Hello! There is a certain array of numbers, for example: the numbers in the arrays are different (from 1 to 52), only 30 numbers in a different sequence. It is necessary to give a forecast for the next step with the probabilities of all 52 numbers falling out. For example: 1 - 10% 2 - 5% ... 52 - 0% Would LSTM work here? And how to properly parse the data?
data = [ [1,2,3,4], [5,6,7,8] ]

Answer 1

First you need to decide if your numbers are random, or they are associated with other numbers in the group, or they form a time series. And if they form, then what should be the dependence - individual in each position in the group, or for the series as a whole.
For random numbers, you can only try to determine the distribution for each position.
For time series, you can try LSTM.