3
$\begingroup$

The Maths behind the problem

Suppose I have a training matrix $X$ with $n$ observations and $d$ features $$ X = \begin{pmatrix} x_{11} & x_{12} & \ldots & x_{1d}\\ x_{21} & x_{22} & \ldots & x_{2d} \\ \vdots & \vdots & \ddots & \vdots \\ x_{n1} & x_{n2} & \ldots & x_{nd} \end{pmatrix} =\begin{pmatrix} {\bf{x}}_1^\top \\ {\bf{x}}_2^\top \\ \vdots \\ {\bf{x}}_n^\top \end{pmatrix} \in\mathbb{R}^{n\times d} $$ Suppose that I want to do a feature transform of this data using the Radial Basis Function. To do this, we

  • choose $b$ rows of $X$ and we call them centroids $$ {\bf{x}}^{(1)}, \ldots, {\bf{x}}^{(b)} $$
  • calculate using some heuristic a bandwidth parameter $\sigma^2$

And then, for every centroid we define a radial basis function as follows $$ \phi^{(i)}({\bf{x}}):=\exp\left(- \frac{\parallel{\bf{x}} - {\bf{x}}^{(i)}\parallel^2}{\sigma^2}\right) \qquad \forall i\in\{1, \ldots, b\} \quad \text{for } {\bf{x}}\in\mathbb{R}^{d} $$ We can therefore obtain a transformed data matrix as $$ \Phi:=\begin{pmatrix} 1 & \phi^{(1)}({\bf{x}}_1) & \phi^{(2)}({\bf{x}}_1) & \cdots & \phi^{(b)}({\bf{x}}_1) \\ 1 & \phi^{(1)}({\bf{x}}_2) & \phi^{(2)}({\bf{x}}_2) & \cdots & \phi^{(b)}({\bf{x}}_2) \\ \vdots & \vdots & \vdots & \ddots & \vdots \\ 1 & \phi^{(1)}({\bf{x}}_n) & \phi^{(2)}({\bf{x}}_n) & \cdots & \phi^{(b)}({\bf{x}}_n) \end{pmatrix} \in\mathbb{R}^{n\times (b+1)} $$ Then we fit a regularized linear model so that optimal parameters are given by $$ {\bf{w}}:=(\Phi^\top\Phi + \lambda I_n)^{-1}\Phi^\top{\bf{y}} $$

Implementation

I have implemented this in both R and Python. I put here my Python implementation.

Parameters

# Set some parameters. Here `reg_coef` is the regularization coefficient
# usually called lambda
n = 1000
d = 2
m = 100
reg_coef = 0.1
b = 2

Create Training Set

To create the response, I simply feed each row of $X$ into a multivariate normal distribution.

# Explanatory variables are uniformly distributed
X = np.random.uniform(-4, 4, size=(n, d))
# Response is a multivariate normal 
target_normal = multivariate_normal(mean=np.random.normal(size=d), cov=np.eye(d))
y = target_normal.pdf(X)

Create test data

Xhat = np.random.uniform(-4, 4, size=(n, d))

Feature Transform

We find $\sigma^2$ with a common heuristic: the median of all the pairwise distances of the data. This step is not important, we could set $\sigma^2$ to pretty much any sensible value.

def compute_sigmasq(X): 
    xcross = np.dot(X, X.T)
    xnorms = np.repeat(np.diag(np.dot(X, X.T)).reshape(1, -1), np.size(X, axis=0), axis=0)
    return(np.median(xnorms - 2*xcross + xnorms.T))
# Find sigmasquared for the rbf
sigmasq = compute_sigmasq(X)

Define a factory of functions for Radial Basis Functions. Basically, given a centroid ${\bf{x}}^{(i)}$, the function rbf returns the function $\phi^{(i)}$ which takes ${\bf{x}}$ as input.

def rbf(centroid):
    # define a closure
    def rbfdot(x):
        return(np.exp(-np.sum((x - centroid)**2) / sigmasq))
    return(rbfdot)

Create a function to compute $\Phi$ the transformed design matrix.

def compute_phiX(X, centroids):
    # Construct phiX
    rbfs = [rbf(centroid) for centroid in centroids]
    list_columns = list(map(lambda f: np.apply_along_axis(f, 1, X), rbfs))
    # Add column on 1s and give correct shape
    list_columns.insert(0, np.repeat(1, np.size(X, axis=0)))
    phiX = np.column_stack(list_columns)
    return(phiX)

Predictions

Define a function that takes a matrix, say $X$, and the number of centroids that we want $b$. It returns the $b$ centoids chosen at random from the rows of the matrix.

def get_centroids(X, n_centroids):
    # Find the indeces
    idx = np.random.randint(np.size(X, axis=0), size=n_centroids)
    # Use indeces to grab rows
    return(X[idx, :])

We also define a function that takes in training data $X$, response vector ${\bf{y}}$, regularization coefficient $\lambda$ and the number of centroids $b$. It then does the whole process of

  • finding the centroids for the training data
  • Computing the matrix $\Phi$ with those new centroids
  • Solving the linear system $(\Phi^\top\Phi + \lambda I_n){\bf{w}} = \Phi^\top {\bf{y}}$ for ${\bf{w}}$

and at the end it returns a prediction function that, given some test data matrix $\widehat{X}\in\mathbb{R}^{m\times d}$ of $m$ new observations, it retuns the predictions $\widehat{y}$

def make_predictor(X, y, reg_coef, n_centroids):
    # Find centroids
    centroids = get_centroids(X, n_centroids)
    # Obtain transformed data matrix
    phiX = compute_phiX(X, centroids)
    # Find optimal parameters
    w = solve(
        np.dot(phiX.T, phiX) + reg_coef*np.eye(n_centroids+1),
        np.dot(phiX.T, y)
    )
    # construct prediction closure
    def predictor(Xhat):
        # Transform test data features
        test_centroids = get_centroids(Xhat, n_centroids)
        phi_Xhat = compute_phiX(Xhat, test_centroids)
        return(np.dot(phi_Xhat, w))
    return(predictor)

To get predictions we then run

# Get predictions and actual values
predict = make_predictor(X, y, reg_coef=reg_coef, n_centroids=b)
yhat = predict(Xhat)
yactual = target_normal.pdf(Xhat)

Notice that we've also found the actual values yhat for the testing set.

Results

I checked the code many times and it looks correct to me. However, if I then plot the predicted values $\widehat{y}$ against the actual values, it looks very weird:

fig, ax = plt.subplots()
ax.scatter(yactual, yhat)
plt.show()

giving

predictions vs real

What am I doing wrong? It looks like they are normally distributed somehow.

EDIT1: Simplified code

The following is a slower version of the code above. It presents the same problem during plotting, however, it is much easier to read.

def compute_phi(X, centroids, sigmasq):
    # X is the matrix to be transformed. b is the number of centroids
    # gen number of samples
    n = X.shape[0]
    b = centroids.shape[0]
    # now slowly construct the matrix by doing a double loop
    values = []
    for x in X:
        for c in centroids:
            values.append(np.exp(-np.sum((x - c)**2) / sigmasq))
    # now simply reshape it
    phiX = np.reshape(values, (n, b))
    return phiX

def predict(Xhat, X, y, b, reg_coef):
    # find centroids and sigmasq
    centroids = get_centroids(X, b)
    sigmasq = compute_sigmasq(X)
    # transform data matrix
    phiX = compute_phi(X, centroids, sigmasq)
    # Find optimal parameters
    w = solve(        
        np.dot(phiX.T, phiX) + reg_coef*np.eye(phiX.shape[1]),
        np.dot(phiX.T, y))
    # Transform test matrix, need to find test centroids
    test_centroids = get_centroids(Xhat, b)
    phi_Xhat = compute_phi(Xhat, test_centroids, sigmasq)
    # Compute predictions with dot product
    return np.dot(phi_Xhat, w)  

The problems arise also if we use

phi_Xhat = compute_phi(Xhat, centroids, sigmasq)

rather than

phi_Xhat = compute_phi(Xhat, test_centroids, sigmasq)
$\endgroup$
12
  • 1
    $\begingroup$ Hi, are you sure, you are allowed to change centroids when predicting? $\endgroup$
    – Konstantin
    Nov 21, 2019 at 15:48
  • $\begingroup$ @Konstantin that is a very valid question. I have no idea, none seems to say it anywhere. I checked Probabilistic Reasoning and Machine Learning by Bishop and also some other resources online, but none specified it. $\endgroup$ Nov 21, 2019 at 15:50
  • $\begingroup$ None seems to tell me whether I should include a column of $1$s in the matrix $\Phi$ either.. $\endgroup$ Nov 21, 2019 at 15:52
  • 1
    $\begingroup$ I am convinced that 1) column of ones is important, and 2) that the centroids are not allowed to shift when you change the inputs (according to the same logic, that makes you keep $w,\sigma$ fixed, they all are parameters of your model, to be determined during training) Check cs.columbia.edu/~jebara/4771/tutorials/regression.pdf $\endgroup$
    – Konstantin
    Nov 21, 2019 at 15:57
  • 1
    $\begingroup$ In compute_phix function, why you have two returns in a row? Also, out of curiosity, why dont use a linear regression package such as glmnet that is perfect for this kind of problems? Also, this line: rbfs = [rbf(centroid) for centroid in centroids] returns None for me. $\endgroup$
    – prony
    Nov 21, 2019 at 16:36

1 Answer 1

2
$\begingroup$

I implemented the problem myself (building on your code) so that you can compare it with your data. I get good results when I use centroids>50. So I think the implementation is correct. I did not use regularization. I simply use the pseudo-inverse function to do regular linear regression

import numpy as np
from scipy.stats import multivariate_normal
import matplotlib.pyplot as plt

np.random.seed(123)


n = 1000
d = 2
m = 100
reg_coef = 0.1
b = 2
n_centroids = 100


# Explanatory variables are uniformly distributed
X = np.random.uniform(-4, 4, size=(n, d))
# Response is a multivariate normal 
target_normal = multivariate_normal(mean=np.random.normal(size=d), cov=np.eye(d))
y = target_normal.pdf(X)

Xhat = np.random.uniform(-4, 4, size=(n, d))
yactual = target_normal.pdf(Xhat)

def compute_sigmasq(X): 
    xcross = np.dot(X, X.T)
    xnorms = np.repeat(np.diag(np.dot(X, X.T)).reshape(1, -1), np.size(X, axis=0), axis=0)
    return(np.median(xnorms - 2*xcross + xnorms.T))
# Find sigmasquared for the rbf
sigmasq = compute_sigmasq(X)

def get_centroids(X, n_centroids):
    # Find the indeces
    idx = np.random.randint(np.size(X, axis=0), size=n_centroids)
    # Use indeces to grab rows
    return(X[idx, :])

centroids = get_centroids(X, n_centroids)


## After this line my implementations #####


def cal_distance(x,cent_i):
    ## calculate the distance between the centroid and the other elements
    return(np.sum((x - cent_i)**2))


phi_train = np.ones((n,1))
for i in range(n_centroids):
    cent_i  = centroids[i,:] # get the ith centroid
    dist_i = np.apply_along_axis(cal_distance , 1, X,cent_i) ## distance matrix for i th centroid
    phi_i = np.exp(-dist_i/sigmasq) ## 
    phi_i = np.reshape(phi_i,(n,1)) # dummy rehsape to (n,1)  
    phi_train = np.hstack((phi_train,phi_i)) # horizontally stack the matrix

inv_phi = np.linalg.pinv(phi_train) # get the psudoinverse for regular linear regression

#linear regression coefficiet
coefs = np.matmul(inv_phi,y) ## compute the weights
coefs = np.reshape(coefs,(n_centroids+1,1))  ## dummy reshape

## create phi test
phi_test = np.ones((n,1))
for i in range(n_centroids):
    cent_i  = centroids[i,:] # get the ith centroid
    dist_i = np.apply_along_axis(cal_distance , 1, Xhat,cent_i) ## distance matrix for i th centroid
    phi_i = np.exp(-dist_i/sigmasq)
    phi_i = np.reshape(phi_i,(n,1)) # dummy rehsape 
    phi_test = np.hstack((phi_test,phi_i))


y_predict = np.matmul(phi_test,coefs)

plt.plot(yactual,y_predict,'*')
plt.xlabel('actual')
plt.ylabel('predicted')
plt.title('n_centroid = ' +  str(n_centroids))

enter image description here enter image description hereenter image description here

This is how the first 10 rows of phi_train look like for n_centroid = 2: enter image description here

Comment

It does not make sense to use different centroids for test data. It is analogous to only use the statistics from training data (e.g., mean, std) while normalizing the data. In theory, you don't have any access to the test data (unseen data). So you are not supposed to make any computation on it (e.g, computing centroids in our case or computing mean for data normalization).

$\endgroup$
5
  • $\begingroup$ I haven't accepted it yet cause I'm still working to understand what went wrong in mine. Even if I set regularization to 0, I still get weird shapes for 100 centroids. Going line by line, I managed to show that we both get the same phi_train, so that's good. Somehow though, our coefficients are different: I think it has to do with the pseudoinverse, but not sure why this happens $\endgroup$ Nov 21, 2019 at 18:51
  • $\begingroup$ I've created the following manual psuedoinverse function def manual_pinv(x): pinv = np.matmul( np.linalg.inv(np.matmul(x.T, x)), x.T) return pinv $\endgroup$ Nov 21, 2019 at 18:52
  • $\begingroup$ This function seems to work for small arrays. For instance, if we create an array some=np.array([[1, 3, 7], [5,1, 2], [4,4,5]]) and we run np.linalg.pinv(some) we get array([[-0.05172414, 0.22413793, -0.01724138], [-0.29310345, -0.39655172, 0.56896552], [ 0.27586207, 0.13793103, -0.24137931]]) and similarly, we get the same if we run manual_pinv(some) $\endgroup$ Nov 21, 2019 at 18:54
  • $\begingroup$ However, if we calculate the pseudo inverse of the phi_train matrix, we get two very different results $\endgroup$ Nov 21, 2019 at 18:54
  • 1
    $\begingroup$ This is a very interesting point. I would also assume that pseudo-inverse equals to $inv(X^TX) X^T$. It turns out, as stated here (docs.scipy.org/doc/numpy/reference/generated/…), the psudo-inverse function (pinv) uses SVD and automatically cuts off for small singular values which consequently regularizes the regression. For example, when I replaced pinv function with $inv(X^TX) X^T$, I still got similar results for n_centroid<50 (slightly worse). However, I got very bad results for n_centroid ~= 100. Because it overfit which was handled automatically by the pinv. $\endgroup$
    – prony
    Nov 21, 2019 at 19:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.