Vectorizing Gradient Descent Function - machine-learning

I'm fairly new to linear algebra and I'm currently taking Andrew Ng's machine learning course. I'm struggling to understand how the two below functions are the same. I'm working on vectorizing gradient descent for linear/logistic regression.
theta = theta - (alpha/m)*(X')*(X*theta - y)
theta = theta - (alpha/m)*sum((X*theta -y)*x(i))
My thought is that x(i) is a vector and in order to do vector multiplication, I need to transpose it, but when trying to mock up an example I didn't see how that was necessary. Any help or explanation would be greatly appreciated.

Assuming you are referring to the equation at the bottom, Lecture 4, slide 8, then you have the sum wrong. The term x(i) is meant to be inside the sum, not outside. And in the 'vectorised' case, your input X contains all individual observations x(i) as (presumably) a column vector (check with your code to be sure). Therefore the correct equivalent lower expression should be:
theta = theta - (alpha/m)*sum((X*theta -y) .* X)
which is indeed equivalent to the other vectorised expression at the top, since in general, for any two (column) vectors a and b, it is true that a.' * b is equivalent to sum(a .* b)

Related

Programmatically performing gradient calculation

Let y = Relu(Wx) where W is a 2d matrix representing a linear transformation on x, a vector. Likewise, let m = Zy, where Z is a 2d matrix representing a linear transformation on y. How do I programmatically calculate the gradient of Loss = sum(m^2) with respect to W, where the power means take the element wise power of the resulting vector, and sum means adding all the elements together?
I can work this out slowly mathematically by taking a hypothetical, multiplying it all out, then element-by-element taking the derivative to construct the gradient, but I can't figure out an efficient approach to write a program once the neural network layer becomes >1.
Say, for just one layer (m = Zy, take gradient wrt Z) I could just say
Loss = sum(m^2)
dLoss/dZ = 2m * y
where * is the outer product of the vectors, and I guess this is kind of like normal calculus and it works. Now for 2 layers + activation (gradient wrt W), if I try to do it like "normal" calculus and apply the chain rule I get:
dLoss/dW = 2m * Z * dRelu * x
where dRelu is the derivative of Relu(Wx) except here I have no idea what * means in this case to make it work.
Is there an easy way to calculate this gradient mathematically without basically multiplying it all out and deriving each separate element in the gradient? I'm really unfamiliar with matrix calculus, so if anyone could also give some mathematical intuition, if my attempt is completely wrong, that would be appreciated.
For the sake of convenience, let's ignore the ReLU for a moment. You have an input space X (of some size [dimX]) mapped to an intermediate space Y (of some size [dimY]) mapped to an output space m (of some size [dimM]) You have, then, W: X → Y a matrix of shape [dimY, dimX] and Z: Y → m a matrix of shape [dimM, dimY]. Finally your loss is simply a function that maps your M space to a scalar value.
Let us walk the way backwards. As you correctly said, you want to compute the derivative of the loss w.r.t W and to do so you need to apply the chain rule all the way back. You then have:
dL/dW = dL/dm * dm/dY * dY/dW
dL/dm is of shape [dimm] (a scalar function with derivatives across dimm dimensions)
dm/dY is of shape [dimm, dimY] (an m-dimensional function with derivatives across dimY dimensions)
dY/dW is of shape [dimY, dimW] = [dimY, dimY, dimX] (a y-dimensional function with derivatives across [dimY, dimX] dimensions)
Edit:
To make the last bit more clear, Y consists of dimY different values, so Y can be treated as dimY constituent functions. We need to apply the gradient operator on each of those mini-functions, all with respect to the basis vectors defined by W. More concretely, if W = [[w11, w12], [w21, w22], [w31, w32]] and x = [x1, x2], then Y = [y1, y2, y3] = [w11x1 + w12x2, w21x1 + w22x2, w31x1 + w32x2]. Then W defines a 6d space (3x2) across which we need to differentiate. We have dY/dW = [dy1/dW, dy2/dW, dy3/dW], and also dy1/dW = [[dy1/dw11, dy1/dw12], [dy1/dw21, dy1/dw22], [dy1/dw31, dy1/dw32]] = [[x1,x2],[0,0],[0,0]], a 3x2 matrix. So dY/dW is a [3,3,2] tensor.
As for the multiplication part; the operation here is tensor contraction (essentially matrix multiplication in high dimension spaces). Practically, if you have a high-order tensor A[[a1, a2, a3... ], β] (i.e. a+1 dimensions, the last of which is of size β) and a tensor B[β, [b1, b2...]] (i.e. b+1 dimensions, the first of which is β), their tensor contraction is a matrix C[[a1,a2...], [b1,b2...]] (i.e. a+b dimensions, the β dimension contracted), where C is obtained by summing over element-wise across the shared dimension β (refer to https://docs.scipy.org/doc/numpy/reference/generated/numpy.tensordot.html#numpy.tensordot).
The resulting tensor contraction then is a matrix of shape [dimY, dimX] which can be used to update your W weights. The ReLU which we ignored earlier can easily be thrown in the mix, since ReLU: 1 → 1 is a scalar function applied element-wise on Y.
To summarize, your code would be:
W_gradient = 2m * np.dot(Z, x) * np.e**x/(1+np.e**x))
I just implemented several multiplier neural networks(MLP) from scratch in C++[1], and I think I know what's your pain. And believe me, you don't even need any third party matrix/tensor/automatic differentiation(AD) libraries to do the matrix multiplication or gradient calculation. There are three things you should pay attention to:
There are two kinds of multiplication in the equations: matrix multiplication, and elementwise multiplication, you'll mess up if you denoted them all as a single *.
Use concrete examples, especially concrete numbers as dimensions of your data/matrix/vector to build intuition.
The most powerful tool for programming correctly is dimension compatibility, always don't forget to check dimensions.
Suppose your want to do binary classification and the neural network is input -> h1 -> sigmoid -> h2 -> sigmoid -> loss in which input layer has 1 sample each has 2 features, h1 has 7 neurons, and h2 has 2 neurons. Then:
forward pass:
Z1(1, 7) = X(1, 2) * W1(2, 7)
A1(1, 7) = sigmoid(Z1(1, 7))
Z2(1, 2) = A1(1, 7) * W2(7, 2)
A2(1, 2) = sigmoid(Z2(1, 2))
Loss = 1/2(A2 - label)^2
backward pass:
dA2(1, 2) = dL/dA2 = A2 - label
dZ2(1, 2) = dL/dZ2 = dA2 * dsigmoid(A2_i) -- element wise
dW2(7, 2) = A1(1, 7).T * dZ2(1, 2) -- matrix multiplication
Notice the last equation, the dimension of the gradient of W2 should match W2, which is (7, 2). And the only way to get a (7, 2) matrix is to transpose input A1 and multiply A1 with dZ2, that's dimension compatibility[2].
backward pass continued:
dA1(1, 7) = dZ2(1, 2) * A1(2, 7) -- matrix multiplication
dZ1(1, 7) = dA1(1, 7) * dsigmoid(A1_i) -- element wise
dW1(2, 7) = X.T(2, 1) * dZ1(1, 7) -- matrix multiplication
[1] The code is here, you can see the hidden layer implementation, naive matrix implementation and the references listed there.
[2] I omit the matrix derivation part, it's simple actually but hard to type the equations out. I strongly suggest you read this paper, every tiny detail you should know on matrix derivation in DL is listed in this paper.
[3] One sample as input is used in the above example(as a vector), you can substitute 1 with any batch numbers(become matrix), and the equations still hold.

how to calculate theta in univariate linear regression model?

I have hypothesis function h(x) = theta0 + theta1*x.
How can I select theta0 and theta1 value for the linear regression model?
The question is unclear whether you would like to do this by hand (with the underlying math), use a program like Excel, or solve in a language like MATLAB or Python.
To start, here is a website offering a summary of the math involved for a univariate calculation: http://www.statisticshowto.com/probability-and-statistics/regression-analysis/find-a-linear-regression-equation/
Here, there is some discussion of the matrix formulation of the multivariate problem (I know you asked for univariate but some people find the matrix formulation helps them conceptualize the problem): https://onlinecourses.science.psu.edu/stat501/node/382
We should start with a bit of an intuition, based on the level of the question. The goal of a linear regression is to find a set of variables, in your case thetas, that minimize the distance between the line formed and the data points observed (often, the square of this distance). You have two "free" variables in the equation you defined. First, theta0: this is the intercept. The intercept is the value of the response variable (h(x)) when the input variable (x) is 0. This visually is the point where the line will cross the y axis. The second variable you have defined is the slope (theta1), this variable expresses how much the response variable changes when the input changes. If theta1 = 0, h(x) does not change when x changes. If theta1 = 1, h(x) increases and decreases at the same rate as x. If theta1 = -1, h(x) responds in the opposite direction: if x increases, h(x) decreases by the same amount; if x decreases, h(x) increases by the quantity.
For more information, Mathworks provides a fairly comprehensive explanation: https://www.mathworks.com/help/symbolic/mupad_ug/univariate-linear-regression.html
So after getting a handle on what we are doing conceptually, lets take a stab at the math. We'll need to calculate the standard deviation of our two variables, x and h(x). WTo calculate the standard deviation, we will calculate the mean of each variable (sum up all the x's and then divide by the number of x's, do the same for h(x)). The standard deviation captures how much a variable differs from its mean. For each x, subtract the mean of x. Sum these differences up and then divide by the number of x's minus 1. Finally, take the square root. This is your standard deviation.
Using this, we can normalize both variables. For x, subtract the mean of x and divide by the standard deviation of x. Do this for h(x) as well. You will now have two lists of normalized numbers.
For each normalized number, multiply the value by its pair (the first normalized x value with its h(x) pair, for all values). Add these products together and divide by N. This gives you the correlation. To get the least squares estimate of theta1, calculate this correlation value times the standard deviation of h(x) divided by the standard deviation of x.
Given all this information, calculating the intercept (theta0) is easy, all we'll have to do is take the mean of h(x) and subtract the product (multiply!) of our calculated theta1 and the average of x.
Phew! All taken care of! We have our least squares solution for those two variables. Let me know if you have any questions! One last excellent resource: https://people.duke.edu/~rnau/mathreg.htm
If you are asking about the hypothesis function in linear regression, then those theta values are selected by an algorithm called gradient descent. This helps in finding the theta values to minimize the cost function.

Analytic gradient of softmax on CS231n

I had a particular question regarding the gradient for the softmax used in the CS231n. After deriving the softmax function to calculate the gradient for each individual class, the authors divide the gradient by the num_examples, even though the gradient is not summed anywhere. What is the logic behind this. Why can't we just use the softmax gradients directly?
A typical objective of a neural network learning is to minimise an expected loss over data distribution, thus:
minimise E_{x,y} L(x,y)
now, in practise we use an estimate of this quantity, which is given by a sample mean, for a training set xi, yi
minimise 1/N SUM L(xi, yi)
what is given in the above derivation is d L(xi, yi) / d theta, but since we want to minimise 1/N SUM L(xi, yi) we should compute its gradient, which is:
d 1/N SUM L(xi, yi) / d theta = 1/N SUM d L(xi, yi) / d theta
This is just a property of partial derivatives (derivative of the sum being a sum of derivatives and so on). Notice, that in all the above derivations author talks about Li, while the actual optimisation is performed over L (notice lack of index i), which is defined as L = 1/N SUM_i Li

Why convolution in spatial domain equal to multiplication in frequency domain?

Why it is said that "convolution of an image in spatial domain is equal to multiplication in frequency domain" ?
Could anyone please explain it briefly?
StackOverflow, unfortunately, doesn't support MathJaX hence it is hard to show the math here.
One way to explain is that Convolution is Linear Invariant Operator.
As you know, Linear Time / Spatially Invariant Systems basically do one thing - Delay and Scaling.
The Eigen Functions of Delay and Scaling are the Harmonic Functions.
Which means that give a signal described by harmonic signals (Practically its Fourier Transform) Linear Time / Spatially Invariant Operator only scales it by complex number (Scaling and shifting by phase) which is what you do in the Fourier Domain.
It is similar to Diagonalization in Linear Algebra.
For instance let's thing of the Filter we apply on the image as an operator - A.
So the output of the system is y = A x.
If A is diagonalizable as A = P^T D P where D is diagonal matrix and P P^T = I, namely Unitary Matrix.
So y = A x = P^T D P x hence by defining z = P x and t = P y we get t = D z namely we only need to multiply each element in t and not the whole matrix multiplication.
If you think about P as the Fourier Transom operator then instead of doing Matrix Multiplication you can have element wise multiplication in other domain - Fourier Domain.

Geometric representation of Perceptrons (Artificial neural networks)

I am taking this course on Neural networks in Coursera by Geoffrey Hinton (not current).
I have a very basic doubt on weight spaces.
https://d396qusza40orc.cloudfront.net/neuralnets/lecture_slides%2Flec2.pdf
Page 18.
If I have a weight vector (bias is 0) as [w1=1,w2=2] and training case as {1,2,-1} and {2,1,1}
where I guess {1,2} and {2,1} are the input vectors. How can it be represented geometrically?
I am unable to visualize it? Why is training case giving a plane which divides the weight space into 2? Could somebody explain this in a coordinate axes of 3 dimensions?
The following is the text from the ppt:
1.Weight-space has one dimension per weight.
2.A point in the space has particular setting for all the weights.
3.Assuming that we have eliminated the threshold each hyperplane could be represented as a hyperplane through the origin.
My doubt is in the third point above. Kindly help me understand.
It's probably easier to explain if you look deeper into the math. Basically what a single layer of a neural net is performing some function on your input vector transforming it into a different vector space.
You don't want to jump right into thinking of this in 3-dimensions. Start smaller, it's easy to make diagrams in 1-2 dimensions, and nearly impossible to draw anything worthwhile in 3 dimensions (unless you're a brilliant artist), and being able to sketch this stuff out is invaluable.
Let's take the simplest case, where you're taking in an input vector of length 2, you have a weight vector of dimension 2x1, which implies an output vector of length one (effectively a scalar)
In this case it's pretty easy to imagine that you've got something of the form:
input = [x, y]
weight = [a, b]
output = ax + by
If we assume that weight = [1, 3], we can see, and hopefully intuit that the response of our perceptron will be something like this:
With the behavior being largely unchanged for different values of the weight vector.
It's easy to imagine then, that if you're constraining your output to a binary space, there is a plane, maybe 0.5 units above the one shown above that constitutes your "decision boundary".
As you move into higher dimensions this becomes harder and harder to visualize, but if you imagine that that plane shown isn't merely a 2-d plane, but an n-d plane or a hyperplane, you can imagine that this same process happens.
Since actually creating the hyperplane requires either the input or output to be fixed, you can think of giving your perceptron a single training value as creating a "fixed" [x,y] value. This can be used to create a hyperplane. Sadly, this cannot be effectively be visualized as 4-d drawings are not really feasible in browser.
Hope that clears things up, let me know if you have more questions.
I have encountered this question on SO while preparing a large article on linear combinations (it's in Russian, https://habrahabr.ru/post/324736/). It has a section on the weight space and I would like to share some thoughts from it.
Let's take a simple case of linearly separable dataset with two classes, red and green:
The illustration above is in the dataspace X, where samples are represented by points and weight coefficients constitutes a line. It could be conveyed by the following formula:
w^T * x + b = 0
But we can rewrite it vice-versa making x component a vector-coefficient and w a vector-variable:
x^T * w + b = 0
because dot product is symmetrical. Now it could be visualized in the weight space the following way:
where red and green lines are the samples and blue point is the weight.
More possible weights are limited to the area below (shown in magenta):
which could be visualized in dataspace X as:
Hope it clarifies dataspace/weightspace correlation a bit. Feel free to ask questions, will be glad to explain in more detail.
The "decision boundary" for a single layer perceptron is a plane (hyper plane)
where n in the image is the weight vector w, in your case w={w1=1,w2=2}=(1,2) and the direction specifies which side is the right side. n is orthogonal (90 degrees) to the plane)
A plane always splits a space into 2 naturally (extend the plane to infinity in each direction)
you can also try to input different value into the perceptron and try to find where the response is zero (only on the decision boundary).
Recommend you read up on linear algebra to understand it better:
https://www.khanacademy.org/math/linear-algebra/vectors_and_spaces
For a perceptron with 1 input & 1 output layer, there can only be 1 LINEAR hyperplane. And since there is no bias, the hyperplane won't be able to shift in an axis and so it will always share the same origin point. However, if there is a bias, they may not share a same point anymore.
I think the reason why a training case can be represented as a hyperplane because...
Let's say
[j,k] is the weight vector and
[m,n] is the training-input
training-output = jm + kn
Given that a training case in this perspective is fixed and the weights varies, the training-input (m, n) becomes the coefficient and the weights (j, k) become the variables.
Just as in any text book where z = ax + by is a plane,
training-output = jm + kn is also a plane defined by training-output, m, and n.
Equation of a plane passing through origin is written in the form:
ax+by+cz=0
If a=1,b=2,c=3;Equation of the plane can be written as:
x+2y+3z=0
So,in the XYZ plane,Equation: x+2y+3z=0
Now,in the weight space;every dimension will represent a weight.So,if the perceptron has 10 weights,Weight space will be 10 dimensional.
Equation of the perceptron: ax+by+cz<=0 ==> Class 0
ax+by+cz>0 ==> Class 1
In this case;a,b & c are the weights.x,y & z are the input features.
In the weight space;a,b & c are the variables(axis).
So,for every training example;for eg: (x,y,z)=(2,3,4);a hyperplane would be formed in the weight space whose equation would be:
2a+3b+4c=0
passing through the origin.
I hope,now,you understand it.
Consider we have 2 weights. So w = [w1, w2]. Suppose we have input x = [x1, x2] = [1, 2]. If you use the weight to do a prediction, you have z = w1*x1 + w2*x2 and prediction y = z > 0 ? 1 : 0.
Suppose the label for the input x is 1. Thus, we hope y = 1, and thus we want z = w1*x1 + w2*x2 > 0. Consider vector multiplication, z = (w ^ T)x. So we want (w ^ T)x > 0. The geometric interpretation of this expression is that the angle between w and x is less than 90 degree. For example, the green vector is a candidate for w that would give the correct prediction of 1 in this case. Actually, any vector that lies on the same side, with respect to the line of w1 + 2 * w2 = 0, as the green vector would give the correct solution. However, if it lies on the other side as the red vector does, then it would give the wrong answer.
However, suppose the label is 0. Then the case would just be the reverse.
The above case gives the intuition understand and just illustrates the 3 points in the lecture slide. The testing case x determines the plane, and depending on the label, the weight vector must lie on one particular side of the plane to give the correct answer.

Resources