Questions on PCA Dimensionality Reduction - math

In machine learning, PCA is used to reduce the dimensionality of training data. However, from the above picture, I can't understand where is the reduction?
The input data x_i has D dimensions:
The output data x still has D dimensions:

The crucial element here is misunderstanding what is the output, in this pseudocode the output is y (equation 29), not x (equation 30), consequently you do reduce your data to d dimensions, the final equation shows you that if you would like to move back to original space, you can do it (obviously data will be recovered with some errors, since in meantime we dropped a lot of information when going to d dimensions).

The important thing to understand while using PCA is the covariance matrix C(x) and its corresponding spectral decomposition. The obtained eigenvalues and eigenvector of the decomposition is used to reduce the dimensionality.
For a D dimensional training set, we have D number of eigenvalues and their corresponding eigenvectors. But in practice (specially image related applications) many of the eigenvectors are correlated; in other words many of them are redundant basis vectors. So discarding those vector from the basis space doesn't result in significant information loss.
Now, if you want to reduce the dimension of your input data from original D to d < D dimension, you can project the input data into d dominant eigenvectors (from the d largest eigenvalues). Eq~29 gives the project input data into the d dimensional space. Eq~30 is used to reconstruct the original data; here reconstruction errors depend on d (number of eigenvectors)

Related

Calculate the reconstruction error as the difference between the original and the reconstructed matrix

I am currently in an online class in genomics, coming in as a wetlab physician, so my statistical knowledge is not the best. Right now we are working on PCA and SVD in R. I got a big matrix:
head(mat)
ALL_GSM330151.CEL ALL_GSM330153.CEL ALL_GSM330154.CEL ALL_GSM330157.CEL ALL_GSM330171.CEL ALL_GSM330174.CEL ALL_GSM330178.CEL ALL_GSM330182.CEL
ENSG00000224137 5.326553 3.512053 3.455480 3.472999 3.639132 3.391880 3.282522 3.682531
ENSG00000153253 6.436815 9.563955 7.186604 2.946697 6.949510 9.095092 3.795587 11.987291
ENSG00000096006 6.943404 8.840839 4.600026 4.735104 4.183136 3.049792 9.736803 3.338362
ENSG00000229807 3.322499 3.263655 3.406379 9.525888 3.595898 9.281170 8.946498 3.473750
ENSG00000138772 7.195113 8.741458 6.109578 5.631912 5.224844 3.260912 8.889246 3.052587
ENSG00000169575 7.853829 10.428492 10.512497 13.041571 10.836815 11.964498 10.786381 11.953912
Those are just the first few columns and rows, it has 60 columns and 1000 rows. Columns are cancer samples, rows are genes
The task is to:
removing the eigenvectors and reconstructing the matrix using SVD, then we need to calculate the reconstruction error as the difference between the original and the reconstructed matrix. HINT: You have to use the svd() function and equalize the eigenvalue to $0$ for the component you want to remove.
I have been all over google, but can't find a way to solve this task, which might be because I don't really get the question itself.
so i performed SVD on my matrix m:
d <- svd(mat)
Which gives me 3 matrices (Eigenassays, Eigenvalues and Eigenvectors), which i can access using d$u and so on.
How do I equalize the eigenvalue and ultimately calculate the error?
https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/svd
the decomposition expresses your matrix mat as a product of 3 matrices
mat = d$u x diag(d$d) x t(d$v)
so first confirm you are able to do the matrix multiplications to get back mat
once you are able to do this, set the last couple of elements of d$d to zero before doing the matrix multiplication
It helps to create a function that handles the singular values.
Here, for instance, is one that zeros out any singular value that is too small compared to the largest singular value:
zap <- function(d, digits = 3) ifelse(d < 10^(-digits) * max(abs(d))), 0, d)
Although mathematically all singular values are guaranteed non-negative, numerical issues with floating point algorithms can--and do--create negative singular values, so I have prophylactically wrapped the singular values in a call to abs.
Apply this function to the diagonal matrix in the SVD of a matrix X and reconstruct the matrix by multiplying the components:
X. <- with(svd(X), u %*% diag(zap(d)) %*% t(v))
There are many ways to assess the reconstruction error. One is the Frobenius norm of the difference,
sqrt(sum((X - X.)^2))

substitution matrix based on spatial autocorrelation transformation

I would like to measure the hamming sequence similarity in which the substitution costs are not based on the substitution rates in the observed sequences but based on the spatial autocorrelation within the study area of the different states (states are thus not related to DNA but something else).
I divided my study area in grid cells of equal size (e.g. 1000m) and measured how often the same "state" is observed in a neighboring cell (Rook-case). Consequently the weight matrix indicates that from state A to A (to move within the same states) has a much higher probability than to go from A to B or B to C or A to C. This already indicates that states have a high spatial autocorrelation.
The problem is, if you want to measure sequence similarity the substitution matrix should be 0 at the diagonal. Therefore I was wondering whether there is a kind of transformation to go from an "autocorrelation matrix" to a substitution matrix, with 0 values along the diagonal. By means of this we would like to account for spatial autocorrelation in the study area in our sequence similarity measure. To do my analysis I am using the package TraMineR.
Example matrix in R for sequences consisting out of four states (A,B,C,D):
Sequence example: AAAAAABBBBCCCCCCCCCCCCDDDDDDDDDDDDDDDDDDDDDDDAAAAAAAAA
Autocorrelation matrix:
A = c(17.50,3.00,1.00,0.05)
B = c(3.00,10.00,2.00,1.00)
C = c(1.00,2.00,30.00,3.00)
D = c(0.05,1.00,3.00,20.00)
subm = rbind(A,B,C,D)
colnames(subm) = c("A","B","C","D")
how to transform this matrix to a substitution matrix?
First, TraMineR computes the Hamming distance, i.e., a dissimilarity, not a similarity.
The simple Hamming distance is just the count of mismatches between two sequences. For example, the Hamming distance between AABBCC and ABBBAC is 2, and between AAAAAA and AAAAAA it is 0 since there are no mismatches.
Generalized Hamming allows to weighting mismatches (not matches!) with substitution costs. For example if the substitution cost between A and B is 1.5, and is 2 between B and C, then the distance would be the weighted sum of mismatches, i.e., 3.5 between the first two sequences. It would still be zero between one sequence and itself.
From what I understand, the shown matrix is not the matrix of substitution costs. It is the matrix of what you call 'spatial autocorrelations', and you look for how you can turn this information into substitutions costs.
The idea is to assign high substitution cost (mismatch weight) when the autocorrelation (a rate in your case) is low, i.e., when there is a low probability to find say state B in the neighborhood of state A, and to assign a low substitution cost when the probability is high. Since your probability matrix is symmetric, a simple solution is to use $1 - p(A|B)$ for all off diagonal terms, and leave 0 on the diagonal for the reason explained above.
sm <- 1 - subm/100
diag(sm) <- 0
sm
For non symmetric probabilities, you could use a similar formula to the one used for deriving the costs from transition rates, i.e., $2 - p(A|B) - p(B|A)$.

fisher's exact test (R) - simulated p-value does not vary

I have a problem using fisher’s exact test in R with a simulated p-value, but I don’t know if it’s a caused by “the technique” ( R ) or if it is (statistically) intended to work that way.
One of the datasets I want to work with:
matrix(c(103,0,2,1,0,0,1,0,3,0,0,3,0,0,0,0,0,0,19,3,57,11,2,87,1,2,0,869,4,2,8,1,4,3,18,16,5,60,60,42,1,1,1,1,21,704,40,759,404,151,1491,9,40,144),ncol=2,nrow=27)
The resulting p-value is always the same, no matter how often I repeat the test:
p = 1 / (B+1)
(B = number of replicates used in the Monte Carlo test)
When I shorten the matrix it works if the number of rows is lower than 19. Nevertheless it is not a matter of number of cells in the matrix. After transforming it into a matrix with 3 columns it still does not work, although it does when using the same numbers in just two columns.
Varying simulated p-values:
>a <- matrix(c(103,0,2,1,0,0,1,0,3,0,0,3,0,0,0,0,0,0,869,4,2,8,1,4,3,18,16,5,60,60,42,1,1,1,1,21),ncol=2,nrow=18)
>b <- matrix(c(103,0,2,1,0,0,1,0,3,0,0,3,0,0,0,0,0,0,19,869,4,2,8,1,4,3,18,16,5,60,60,42,1,1,1,1,21,704),ncol=2,nrow=19)
>c <- matrix(c(103,0,2,1,0,0,1,0,3,0,0,3,0,0,0,0,0,0,869,4,2,8,1,4,3,18,16,5,60,60,42,1,1,1,1,21),ncol=3,nrow=12)
>fisher.test(a,simulate.p.value=TRUE)$p.value
Number of cells in a and b are the same, but the simulation only works with matrix a.
Does anyone know if it is a statistical issue or a R issue and, if so, how it could be solved?
Thanks for your suggestions
I think that you are just seeing a very significant result. The p-value is being computed as the number of simulated (and the original) matrices that are as extreme or more extreme than the original. If none of the randomly generated matrices are as or more extreme then the p-value will just be 1 (the original matrix is as extreme as itself) divided by the total number of matrices which is $B+1$ (the B simulated and the 1 original matrix). If you run the function with enough samples (high enough B) then you will start to see some of the random matrices as or more extreme and therefor varying p-values, but the time to do so is probably not reasonable.

An "asymmetric" pairwise distance matrix

Suppose there are three sequences to be compared: a, b, and c. Traditionally, the resulting 3-by-3 pairwise distance matrix is symmetric, indicating that the distance from a to b is equal to the distance from b to a.
I am wondering if TraMineR provides some way to produce an asymmetric pairwise distance matrix.
No, TraMineR does not produce 'assymetric' dissimilaries precisely for the reasons stressed in Pat's comment.
The main interest of computing pairwise dissimilarities between sequences is that once we have such dissimilarities we can for instance
measure the discrepancy among sequences, determine neighborhoods, find medoids, ...
run cluster algorithms, self-organizing maps, MDS, ...
make ANOVA-like analysis of the sequences
grow regression trees for the sequences
Inputting a non symmetric dissimilarity matrix in those processes would most probably generate irrelevant outcomes.
It is because of this symmetry requirement that the substitution costs used for computing Optimal Matching distances MUST be symmetrical. It is important to not interpret substitution costs as the cost of switching from one state to the other, but to understand them for what they are, i.e., edit costs. When comparing two sequences, for example
aabcc and aadcc, we can make them equal either by replacing arbitrarily b with d in the first one or d with b in the second one. It would then not make sense not giving the same cost for the two substitutions.
Hope this helps.

What is SVD(singular value decomposition)

How does it actually reduce noise..can you suggest some nice tutorials?
SVD can be understood from a geometric sense for square matrices as a transformation on a vector.
Consider a square n x n matrix M multiplying a vector v to produce an output vector w:
w = M*v
The singular value decomposition M is the product of three matrices M=U*S*V, so w=U*S*V*v. U and V are orthonormal matrices. From a geometric transformation point of view (acting upon a vector by multiplying it), they are combinations of rotations and reflections that do not change the length of the vector they are multiplying. S is a diagonal matrix which represents scaling or squashing with different scaling factors (the diagonal terms) along each of the n axes.
So the effect of left-multiplying a vector v by a matrix M is to rotate/reflect v by M's orthonormal factor V, then scale/squash the result by a diagonal factor S, then rotate/reflect the result by M's orthonormal factor U.
One reason SVD is desirable from a numerical standpoint is that multiplication by orthonormal matrices is an invertible and extremely stable operation (condition number is 1). SVD captures any ill-conditioned-ness in the diagonal scaling matrix S.
One way to use SVD to reduce noise is to do the decomposition, set components that are near zero to be exactly zero, then re-compose.
Here's an online tutorial on SVD.
You might want to take a look at Numerical Recipes.
Singular value decomposition is a method for taking an nxm matrix M and "decomposing" it into three matrices such that M=USV. S is a diagonal square (the only nonzero entries are on the diagonal from top-left to bottom-right) matrix containing the "singular values" of M. U and V are orthogonal, which leads to the geometric understanding of SVD, but that isn't necessary for noise reduction.
With M=USV, we still have the original matrix M with all its noise intact. However, if we only keep the k largest singular values (which is easy, since many SVD algorithms compute a decomposition where the entries of S are sorted in nonincreasing order), then we have an approximation of the original matrix. This works because we assume that the small values are the noise, and that the more significant patterns in the data will be expressed through the vectors associated with larger singular values.
In fact, the resulting approximation is the most accurate rank-k approximation of the original matrix (has the least squared error).
To answer to the tittle question: SVD is a generalization of eigenvalues/eigenvectors to non-square matrices.
Say,
$X \in N \times p$, then the SVD decomposition of X yields X=UDV^T where D is diagonal and U and V are orthogonal matrices.
Now X^TX is a square matrice, and the SVD decomposition of X^TX=VD^2V where V is equivalent to the eigenvectors of X^TX and D^2 contains the eigenvalues of X^TX.
SVD can also be used to greatly ease global (i.e. to all observations simultaneously) fitting of an arbitrary model (expressed in an formula) to data (with respect to two variables and expressed in a matrix).
For example, data matrix A = D * MT where D represents the possible states of a system and M represents its evolution wrt some variable (e.g. time).
By SVD, A(x,y) = U(x) * S * VT(y) and therefore D * MT = U * S * VT
then D = U * S * VT * MT+ where the "+" indicates a pseudoinverse.
One can then take a mathematical model for the evolution and fit it to the columns of V, each of which are a linear combination the components of the model (this is easy, as each column is a 1D curve). This obtains model parameters which generate M? (the ? indicates it is based on fitting).
M * M?+ * V = V? which allows residuals R * S2 = V - V? to be minimized, thus determining D and M.
Pretty cool, eh?
The columns of U and V can also be inspected to glean information about the data; for example each inflection point in the columns of V typically indicates a different component of the model.
Finally, and actually addressing your question, it is import to note that although each successive singular value (element of the diagonal matrix S) with its attendant vectors U and V does have lower signal to noise, the separation of the components of the model in these "less important" vectors is actually more pronounced. In other words, if the data is described by a bunch of state changes that follow a sum of exponentials or whatever, the relative weights of each exponential get closer together in the smaller singular values. In other other words the later singular values have vectors which are less smooth (noisier) but in which the change represented by each component are more distinct.

Resources