5
1 Multivariate Data Analysis (MVDA) Any investigation of a real process or system is based on data measurements The data collected in science, technology and almost everywhere else are multivariate; with multiple variables measured on multiple samples or at multiple time points cannot be analyzed by simple graphs More sophisticated, computer based methods such as Principle Compenent Analysis are required for a multivariate data set New information and facts can be seen using MVDA PCA is one of the most popular MVDA technique Convert data tables to plots

Multivariate Data Analysis (MVDA) - QualityNano Data Analysis (MVDA) •Any investigation of a real process or system is based on data ... Principal Component Analysis (PCA) Overview

Embed Size (px)

Citation preview

Page 1: Multivariate Data Analysis (MVDA) - QualityNano Data Analysis (MVDA) •Any investigation of a real process or system is based on data ... Principal Component Analysis (PCA) Overview

1

Multivariate Data Analysis (MVDA)•Any investigation of a real process or system is based on datameasurements•The data collected in science, technology and almost everywhere elseare multivariate; with multiple variables measured on multiple samplesor at multiple time points cannot be analyzed by simple graphs•More sophisticated, computer based methods such as PrincipleCompenent Analysis are required for a multivariate data set•New information and facts can be seen using MVDA•PCA is one of the most popular MVDA technique

Convert data tables to plots

Page 2: Multivariate Data Analysis (MVDA) - QualityNano Data Analysis (MVDA) •Any investigation of a real process or system is based on data ... Principal Component Analysis (PCA) Overview

2

with 104 observations, K = 34 X variables,and M = 1 Y‐variable

X3

X1

X2

X Y104 104

K=34 M=1

observations

factors response

Provides an overview of therelationships among all X‐variables and Y variables at the same time.

PLS

Page 3: Multivariate Data Analysis (MVDA) - QualityNano Data Analysis (MVDA) •Any investigation of a real process or system is based on data ... Principal Component Analysis (PCA) Overview

3

Projection methods (PCA and PLS)

X X Y

Overview andSummary

• PCA, PrincipalComponents Analysis

Relation between blocksof variables, X & Y• PLS analysis• PLS‐DA

• PCA models the correlationstructure of a data set

Partial Least Squares Projections to Latent Structures (PLS)‐Relating X to Y

Principal Component Analysis (PCA)Overview of data tables

• PLS find relationships betweensets of multivariate data X and Y

PLS differences to PCA*Projection of X that  *Projection of X thatis an optimal both

approximation of X approximates X well, (least squares fit) and correlates well with Y 

Page 4: Multivariate Data Analysis (MVDA) - QualityNano Data Analysis (MVDA) •Any investigation of a real process or system is based on data ... Principal Component Analysis (PCA) Overview

4

Case Study 2Gini (1999) used a  back propagation neural network to predict the

carcinogenicity of aromatic Nitrogen compounds.34 Descriptors 104 Molecules

Carcinogenicity

Y

Y=f(x)

Molecular Descriptors

X

QSAR

Gini. G. Et al. (1999). Predictive carcinogenicity: A model for aromatic compounds with nitrogen-containing substituents, based on moleculardescriptors using an artifical neural network. J.of Che. Inf. Com. Sci., 1076-1080

Page 5: Multivariate Data Analysis (MVDA) - QualityNano Data Analysis (MVDA) •Any investigation of a real process or system is based on data ... Principal Component Analysis (PCA) Overview

5

Variables close to each other correlateObservations close to each other are similar.