Robust Methods in Analysis of Multivariate Food Chemistry Data

摘要: Abstract In this chapter, foundations of robust statistics are introduced, including classic and estimators as well their statistical properties (breakdown point, efficiency, influence function equivariance property). Then, some methods that have gained popularity in recent years presented. The major benefit using stems from the fact they help providing stable estimates for data containing outliers (food samples considerably different compositions comparison with majority samples). Regardless reasons uniqueness, strongly affect interpretation when any method least-squares cost is used. Therefore, more suitable to explore model natural expected. addition exploration modeling multivariate data, processing incomplete contain also discussed.