Noise

And let's start with a basic problem.

Noise makes it hard to answer three questions.

When did a change start? When did the process settle? And what is the true value right now? So in his articles, Russ opens up with a simulated example.

The process moves from one steady value to another, and you see noisy data points with the first order filtered down and drawn through them.

That way, proportional action stays responsive to the actual data instead of the lagged version of it.

So as I said earlier, hang on because part two changes a little bit more, and that's what we're gonna talk about because not all noise is random.

So in part two, it was titled, Not All Processed Noise is Random.

This is where the ideal assumptions get challenged.