7  Principal Components Analysis

SVD
PCA

Overview

Principal Component Analysis (PCA) is a method of dimension reduction. This is not directly related to prediction problem, but several regression methods are directly dependant on it. The regression methods (PCR and PLS) will be considered later. Now a motivation for dimension reduction is being set up.

Notation

The input matrix X of dimension \(N \times p\):

\[\begin{pmatrix} x_{1,1} & x_{1,2} & ... & x_{1,p} \\ x_{2,1} & x_{2,2} & ... & x_{2,p}\\ ... & ... & ... & ...\\ x_{N,1} & x_{N,2} & ... & x_{N,p} \end{pmatrix}\]

The rows of the above matrix represent the cases or observations.

The columns represent the variables observed in each unit. These represent the characteristics.

Assume that the columns of X are centered, i.e., the estimated column mean is subtracted from each column.

Objectives

Upon successful completion of this lesson, you should be able to:


  • Introducing Principal Component Analysis.
  • The precursor to Regression Techniques after Dimension Reduction.

7.1 Singular Value Decomposition (SVD)

Singular Value Decomposition (SVD)

Singular value decomposition is the key part of principal components analysis.

The SVD of the \(N × p\) matrix \(\mathbf{X}\) has the form \(\mathbf{X} = \mathbf{U}\mathbf{D}\mathbf{V}^T\).

  • \(\mathbf{U} = (\mathbf{u}_1, \mathbf{u}_2, \cdots , \mathbf{u}_N)\) is an N × N orthogonal matrix. \(\mathbf{u}_j, j = 1, \cdots , N\), form an orthonormal basis for the space spanned by the column vectors of \(\mathbf{X}\).
  • \(\mathbf{V} = (\mathbf{v}_1, \mathbf{v}_2, \cdots , \mathbf{v}_p)\) is an p × p orthogonal matrix. \(\mathbf{v}_j, j = 1, \cdots , p\), form an orthonormal basis for the space spanned by the row vectors of \(\mathbf{X}\).
  • \(\mathbf{D}\) is a N x p rectangular matrix with nonzero elements along the first p x p submatrix diagonal. \(diag(\mathbf{d}_1, \mathbf{d}_2, \cdots , \mathbf{d}_p)\), \(\mathbf{d}\_1 \geq \mathbf{d}\_2 \geq \cdots \geq \mathbf{d}\_p \geq 0\) are the singular values of \(\mathbf{X}\) with N > p.

The columns of \(\mathbf{V}\) (i.e., \(\mathbf{v}_j, j = 1, \cdots , p)\) are the eigenvectors of \(\mathbf{X}^T\mathbf{X}\). They are called principal component direction of \(\mathbf{X}\).

The diagonal values in \(\mathbf{D}\) (i.e., \(\mathbf{d}_j, j = 1, \cdots , p)\) are the square roots of the eigenvalues of \(\mathbf{X}^T\mathbf{X}\).

The Two-Dimensional Projection

The two-dimensional plane can be shown to be spanned by

  1. the linear combination of the variables that has maximum sample variance,
  2. the linear combination that has maximum variance subject to being uncorrelated with the first linear combination.

It can be extended to the k-dimensional projection. We can take the process further, seeking additional linear combinations that maximize the variance subject to being uncorrelated with all those already selected.

7.2 Principal Components

Principal components analysis is one of the most common methods used for linear dimension reduction. The motivation behind dimension reduction is that the process gets unwieldy with a large number of variables while the large number does not add any new information to the process. An analogy may be drawn with variance inflation factors in multiple regression. If VIF corresponding to any predictor is large, that predictor is not included in the model, as that variable does not contribute any new information. On the other hand, because of linear dependence, the regression matrix may become singular. In a multivariate situation, it may well happen that, a few (or a large number of) variables have high interdependence. A linear combination of variables is then considered which are orthogonal to one another, but the total variability within the sample is preserved as much as possible.

Suppose the data is 10-dimensional but needs to be reduced to 2-dimensional. The idea of principal component analysis is to use two directions that capture the variation in the data as much as possible.

[Keep in mind that the dimensions to which data needs to be reduced are usually not pre-fixed. After taking a look at the total proportion of variability captured, the reduction in dimension is determined. Here reduction of 10-dimensional space to 2-dimension space is for illustration only]

The sample covariance matrix of X is given as:

\[S=\textrm{X}^T\textrm{X} /N\]

If you do the Eigen decomposition of \(X^TX\):

\[\begin{align*}\textrm{X}^T\textrm{X} &= (\textrm{U}\textrm{D}\textrm{V}^T)^T(\textrm{U}\textrm{D}\textrm{V}^T)\\ &= \textrm{V}\textrm{D}^T\textrm{U}^T\textrm{U}\textrm{D}\textrm{V}^T\\ &=\textrm{V}\textrm{D}^2\textrm{V}^T \end{align*}\]

It turns out that if you have done the singular value decomposition then you already have the Eigen value decomposition for \(X^TX\).

The \(D^2\) is the diagonal part of matrix D with every element on the diagonal squared.

Also, we should point out that we can show using linear algebra that \(X^TX\) is a semi-positive definite matrix. This means that all of the eigenvalues are guaranteed to be nonnegative. The eigen values are in matrix \(D^2\). Since these values are squared, every diagonal element is non-negative.

The eigenvectors of \(X^TX\), \(v_j\), can be obtained either by doing an Eigen decomposition of \(X^TX\), or by doing a singular value decomposition from X. These vj’s are called principal component directions of X. If you project X onto the principal components directions you get the principal components.

\[ \textrm{z}_1=\begin{pmatrix} X_{1,1}v_{1,1}+X_{1,2}v_{1,2}+ ... +X_{1,p}v_{1,p}\\ X_{2,1}v_{1,1}+X_{2,2}v_{1,2}+ ... +X_{2,p}v_{1,p}\\ \vdots\\ X_{N,1}v_{1,1}+X_{N,2}v_{1,2}+ ... +X_{N,p}v_{1,p}\end{pmatrix}\]

  • It’s easy to see that \(\mathbf{z}_j = \mathbf{X}\mathbf{v}_j = \mathbf{u}_j d_j\) . Hence \(\mathbf{u}_j\) is simply the projection of the row vectors of \(\mathbf{X}\), i.e., the input predictor vectors, on the direction \(\mathbf{v}_j\), scaled by \(d_j\). For example:
  • The principal components of \(\mathbf{X}\) are \(\mathbf{z}_j = d_j \mathbf{u}_j , j = 1, ... , p\) .
  • The first principal component of \(\mathbf{X}\), \(\mathbf{z}_1\) , has the largest sample variance amongst all normalized linear combinations of the columns of \(\mathbf{X}\).
  • Subsequent principal components \(\mathbf{z}_j\) have maximum variance \(d^2_j / N\), subject to being orthogonal to the earlier ones.
Geometric graph of the PCA analysis

View the Video Explanation

PCA Graph

Why are we interested in this? Consider an extreme case, (lower right), where your data all lie in one direction. Although two features represent the data, we can reduce the dimension of the dataset to one using a single linear combination of the features (as given by the first principal component).

Just to summarize the way you do dimension reduction. First you have X, and you remove the means. On X you do a singular value decomposition and obtain matrices U, D, and V. Then you call every column vector in V, \(v_j\), where j = 1, … , p. This vj is called the principal components direction. Let’s say your original dimension is 10 and we wish to reduce this to 2 dimensions, what would we do? We would take \(v_1\) and \(v_2\) and project X to \(v_1\) and X to \(v_2\). To project, multiply X by \(v_1\) for a column vector of size N. X times \(v_2\) gives you another column vector of size N. These two column vectors reduce the dimensions so you have:

\[\begin{pmatrix} \tilde{x}_{1,1} & \rightarrow & \tilde{x}_{1,v} \\ \tilde{x}_{2,1} & \rightarrow & \tilde{x}_{1,v} \\ \downarrow & & \downarrow\\ \tilde{x}_{N,1} & \rightarrow & \tilde{x}_{N,v} \end{pmatrix}\]

Every row would be a reduced version of your original data. If I asked you to give me the 2 dimensional version of the second sample point, then you would you would look at the row highlighted below:

7.3 Principal Components Analysis (PCA)

Objective

Capture the intrinsic variability in the data.

Reduce the dimensionality of a data set, either to ease interpretation or as a way to avoid overfitting and to prepare for subsequent analysis.

The sample covariance matrix of \(\mathbf{X}\) is \(\mathbf{S} = \mathbf{X}^T\mathbf{X}/\mathbf{N}\), since \(\mathbf{X}\) has zero mean.

Eigen decomposition of \(\mathbf{X}^T\mathbf{X}\):

\[\mathbf{X}^T\mathbf{X} = (\mathbf{U}\mathbf{D}\mathbf{V}^T)^T (\mathbf{U}\mathbf{D}\mathbf{V}^T) =\mathbf{V}\mathbf{D}^T\mathbf{U}^T\mathbf{U}\mathbf{D}\mathbf{V}^T = \mathbf{V}\mathbf{D}^2\mathbf{V}^T\]

The eigenvectors of \(\mathbf{X}^T\mathbf{X}\) (i.e.,\(v _ { j } j = 1 , \dots , p\) ) are called principal component directions of \(\mathbf{X}\).

The first principal component direction \(\mathbf{v}_1\) has the following properties that

  • \(\mathbf{v}_1\) is the eigenvector associated with the largest eigenvalue, \(\mathbf{d}_1^2\), of \(\mathbf{X}^T\mathbf{X}\).
  • \(\mathbf{z}_1 = \mathbf{X}\mathbf{v}_1\) has the largest sample variance amongst all normalized linear combinations of the columns of X.
  • \(\mathbf{z}_1\) is called the first principal component of \(\mathbf{X}\). And, we have \(Var(\mathbf{z}_1)= d_1^2 / N\).

The second principal component direction \(v_2\) (the direction orthogonal to the first component that has the largest projected variance) is the eigenvector corresponding to the second largest eigenvalue, \(\mathbf{d}_2^2\), of \(\mathbf{X}^T\mathbf{X}\), and so on. (The eigenvector for the \(k^{th}\) largest eigenvalue corresponds to the \(k^{th}\) principal component direction \(\mathbf{v}_k\).)

The \(k^{th}\) principal component of \(\mathbf{X}\), \(\mathbf{z}_k\), has maximum variance \(\mathbf{d}_1^2 / N\), subject to being orthogonal to the earlier ones.

Geometric graph of the PCA analysis

7.4 Geometric Interpretation

Principal components analysis (PCA) projects the data along the directions where the data varies the most.

PCA plot

The first direction is decided by \(\mathbf{v}_1\) corresponding to the largest eigenvalue \(d_1^2\).

The second direction is decided by \(\mathbf{v}_2\) corresponding to the second largest eigenvalue \(d_2^2\).

The variance of the data along the principal component directions is associated with the magnitude of the eigenvalues.

Choice of How Many Components to Extract

Scree Plot – This is a useful visual aid which shows the amount of variance explained by each consecutive eigenvalue.

The choice of how many components to extract is fairly arbitrary.

When conducting principal components analysis prior to further analyses, it is risky to choose too small a number of components, which may fail to explain enough of the variability in the data.


7.5 R Scripts

1. Acquire Data

Diabetes data

The diabetes data set is taken from the UCI machine learning database on Kaggle: Pima Indians Diabetes Database

  • 768 samples in the dataset
  • 8 quantitative variables
  • 2 classes; with or without signs of diabetes

Save the data into your working directory for this course as “diabetes.data.” Then load data into R as follows:

# comma delimited data and no header for each variable
RawData = read.table("diabetes.data",sep =",",header=FALSE)

In RawData, the response variable is its last column; and the remaining columns are the predictor variables.

responseY = RawData[,dim(RawData)[2]]
predictorX = RawData[,1:(dim(RawData)[2]-1)]

2. Principal Component Analysis

Principal component analysis is done by the princomp function. For computing the principal components, sometimes it is recommended the data be scaled first. However, one needs to judge whether scaling is necessary on a case by case base. For scaling, we can set the cor=T argument.

pca = princomp(predictorX, cor=T) # principal components analysis using correlation matrix

$scores gives the principal components arranged in decreasing order of the standard deviations of the principal components. Suppose we want to get the first and second principal components. We use the following code. The scatter plot for the two components is then drawn.

pc.comp = pca$scores
    pc.comp1 = -1*pc.comp[,1] # principal component 1 scores (negated for convenience)
    pc.comp2 = -1*pc.comp[,2] # principal component 2 scores (negated for convenience)
    plot(pc.comp1, pc.comp2, xlim=c(-6,6), ylim=c(-3,4), type="n")
    points(pc.comp1[responseY==0], pc.comp2[responseY==0], cex=0.5, col="blue")
    points(pc.comp1[responseY==1], pc.comp2[responseY==1], cex=0.5, col="red")

The scattor plot of the first two principal components are shown in the following figure. The red circles show data in Class 1 (cases with diabetes), and the blue circles show Class 0 (non diabetes).

R output
Figure 1: The scatter plot of the first two principal components for the Diabetes data

7.6 More Examples

Example 1: Handwritten Digit Recognition

  • Goal: Identify single digits 0 \(\sim\) 9 based on images.
  • Raw data: Images that are scaled segments from five digit ZIP codes.
  • \(16\times16\) eight-bit grayscale maps
  • Pixel intensities range from 0 (black) to 255 (white)
  • Input data: represent each image as a high-dimensional vector \(x \in \mathbb{R}^{256}\).

PCA can help you to transform the high dimension image data into lower dimension principal components.

Example: Face Recognition

The cumulative effect of nine principal components, adding one PC at a time, for “sad”. The more principal components we use the better resolution we get. However, 4 or 5 principal components lead to a good judgment on a sad expression. It is a dramatic dimension reduction considering the original number of variables which is the number of pixels for a figure.

Panel of 9 faces with different levels of pixelation

Why do dimensionality reduction?

  • Computational: compress data \(\Rightarrow\) time/space efficiency.
  • Statistical: fewer dimensions \(\Rightarrow\) better generalization.
  • Visualization: understand the structure of data.
  • Anomaly detection: describe normal data, detect outliers.

When faced with situations involving high-dimensional data, it is natural to consider projecting those data onto a lower-dimensional subspace without losing important information.

  • Variable selection also called feature selection.
  • Shrinkage: Ridge regression and Lasso.
  • Creating a reduced set of linear or nonlinear transformations of the input variables, also called feature extraction, e.g. PCA.