How to check data integrity using neural networks?

Question

Neural networks

How can you check the integrity of the data of technological protocols (for example, readings from sensors) using neural networks?
There is an idea to make a data classifier (broken data / integer data), but I got stuck on the issue of determining the data bitness.

Answer 1

I don’t understand what side the neural networks are here. But if they are considered as approximating functions, then they may well fit.
To get started, read probability theory, mathematical statistics, and hypothesis testing.
Next, try to comprehend the application of neural networks in these areas.

Answer 2

Maybe it was meant to detect outliers and anomalies in time series, which represent parameters that change over time? Then the task becomes meaningful. True, from my point of view, neural networks are not the best solution here, there are others that are more adequate, but nevertheless, as a project, it can work.