Dealing with missing values and outliers in principal component analysis

Stanimirova, I.; Daszykowski, M.; Walczak, Beata

doi:10.1016/j.talanta.2006.10.011

Cited by 112 publications

(56 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…PCA and TUCKER3 methods are often used in exploratory analysis of environmental monitoring data (Stanimirova et al 2006(Stanimirova et al , 2007. Unfortunately, with these methods only complete data can be analyzed effectively.…”

Section: Resultsmentioning

confidence: 99%

“…This paper presents a way of ameliorating the problem of missing elements using the expectation-maximization (EM) algorithm (McLaachlan & Krishnan 1997;Walczak & Massart 2001a,b;Smoliński & Walczak 2002;Stanimirova et al 2007), which may be built into different computational procedures such as EM/PCA or EM/TUCKER3. Problems associated with exploratory analyses of data sets with missing elements are demonstrated on a real environmental data set, that of Gdańsk Deep water column chemical component concentrations (Pryputniewicz 2007).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Chemometric exploration of sea water chemical component data sets with missing elements

Smoliński

Falkowska

Pryputniewicz

2008

Oceanological and Hydrobiological Studies

View full text Add to dashboard Cite

The results of the application of chemometric methods, such as principal component analysis (PCA) and its generalization for N-way data, the Tucker3 model, for the analysis of an environmental data set are presented. The analyzed data consists of concentration values of chemical compounds of organic matter, and their transformed products, in a short-term study of a sea water column measured at the Gdańsk Deep (φ = 55°1'N, λ = 19°10'E).The main goal of this paper is to present improved approaches for exploration of data sets with missing elements, based on the expectation-maximization (EM) algorithm. The most common methods for dealing with missing data, generally consisting of setting the missing elements to zero or to mean values of the measured data, are often unacceptable as they destroy data correlations or influence interpretation of relationships between objects and variables. The EM algorithm may be built into different computational procedures used for exploratory analysis (i.e. EM/PCA or EM/TUCKER3).

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Chemometric exploration of sea water chemical component data sets with missing elements

Smoliński

Falkowska

Pryputniewicz

2008

Oceanological and Hydrobiological Studies

View full text Add to dashboard Cite

show abstract

“…PCs are actually orthogonal and linear combinations of the data for their maximum diversity. Simultaneously, each of the PCs carries different information about data variability [3,7,34]. PCA allows us to group the experimental data and present them in the form of several PCs, and PCs contain nearly identical information to the experimental data.…”

Section: Methodsmentioning

confidence: 99%

“…Thus, the complete input data are obtained for PCA. The distribution of values in the full dataset and the prediction of X in the basic model of PCA are executed [3][4][5]:…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

The Possibility of Applying the EM-PCA Procedure to Lake Water

Bucior-Kwaczyńska¹

2018

Pol. J. Environ. Stud.

View full text Add to dashboard Cite

A new fault detection method based on artificial immune systems

Wang¹,

Zhao²

2008

Asia-Pacific J Chem Eng

View full text Add to dashboard Cite

A new fault detection method with a continuous learning feature for a complicated process is proposed based on the concept of artificial immune systems (AIS). Both the negative and the clonal selections are adopted in the method. The real-valued negative selection algorithm (RNSA) is utilized to generate fault detectors. When the detector set is used to perform the fault detection, a clonal selection is employed to update the fault detector set. The proposed method is applied to the Tennessee Eastman (TE) process. The simulation results show that the performance of the proposed method is superior to those of both classical principal component analysis (PCA) and negative selection algorithm. 

show abstract

Dealing with missing values and outliers in principal component analysis

Cited by 112 publications

References 16 publications

Chemometric exploration of sea water chemical component data sets with missing elements

Chemometric exploration of sea water chemical component data sets with missing elements

The Possibility of Applying the EM-PCA Procedure to Lake Water

A new fault detection method based on artificial immune systems

Contact Info

Product

Resources

About