Hey Resuly, looking at this just now, I can see why this is confusing, I'll see how to word this better later on.

Anyway, what the text is trying to say is that:

target_value = observed_value and,

observed_value = actual_value + Error

Further, the error is normally distributed.

If you step back and don't look and this from a mathematical perspective, the thing to realize is that in real world experiments all you have are your observed_values which have some margin of error, those observed_values become your features and targets.

🍻