Energy-biased detection probability
= Energy-biased detection probability
{title2=$P(i\mid D)=p_i\eta_i/\sum_jp_j\eta_j$}
If source class $i$ has prior fraction $p_i$ and <detection efficiency> $\eta_i$, its fraction among detected events is proportional to $p_i\eta_i$. <Bayes' theorem> normalizes these products. Class-dependent losses therefore bias the detected composition even when the incident fractions are known exactly.