Regression

The (controlled) trial is a common paradigm from experimental sciences. Under experimental conditions x some measurement result y is recorded. This is composed of a systematic effect m(x) and a measurement error ε:

y = m(x) + ε.