Eemoutlier

Purpose

Automatically remove outliers in PARAFAC models of EEM data.

Synopsis

result = eemoutlier(X,factors);

Description

Provides an automated outlier selection procedure for a given dataset and given number of components. Samples with a high leverage or high sum-squared residual are removed one by one until no samples are assessed as outliers. The settings for making decisions are given in the top of the m-file. Outliers can only be removed until there are 8 samples left. Then the algorithm will stop.

Inputs

• X = multi-way array.
• factors = number of factors.

Outputs

• result = structure with following fields:

result.model = parafac model.

result.SMPS = cell of sample sets.

See Also

parafac