I would like to simulate the noise or uncertainty present in spectrophotometric absorbance readings used for quantifying cell counts using cell viability assays. Which is more suitable, additive or multiplicative noise? Or if there exist other type of noise modeling that is more suitable than either?
What I mean by these two types of (Gaussian) noise is:
- Additive noise: y = ytrue + N(0, var)
- Multiplicative noise: y = y * exp(N(0, var))
*Another one which I encountered before is y = y + y*N(0, var)
My intuition is that additive noise is probably more suitable because for multiplicative noise, the noise scales with y value, which does not seem like a reasonable assumption for absorbance reading.
- Badges
- Science topic
- Similar topics
- Mathematics
More Tianxun Zhou's questions See All
Similar questions and discussions