Iterative Log Mean Centering Normalization

Parameter Information


Standard Deviation

Iterative Log Mean Centering Normalization requires only a standard deviation range to be selected. During each iteration the a mean log ratio and standard deviation is taken on a subset of the data which is initially the entire data set. The calculated mean is used to correct the position to move the mean toward zero. At the end of each iteration, data elements within the indicated standard deviation range are selected to enter the next iteration and are used to calculate a current mean and standard deviation. Iterations stop when the mean converges to zero.

Using small values for standard deviation will force the subset of data used for the next iteration to be a more aggressive cut of the data.