梯度下降（Gradient Descent）

我們將使用梯度下降算法來(lái)求出使得代價(jià)函數(shù)J(θ₀, θ₁)值最小的參數(shù)θ的值公条。

梯度下降算法的基本思想：首先我們隨機(jī)選擇一個(gè)參數(shù)的組合(θ₀, θ₁, ... , θ_n)拇囊，計(jì)算代價(jià)函數(shù)，然后我們尋找下一個(gè)能讓代價(jià)函數(shù)值最小的參數(shù)組合靶橱，且一直這樣尋找下去直至找到一個(gè)局部最小值寥袭。為什么將這個(gè)最小值稱(chēng)為局部最小值？因?yàn)楣匕裕覀冊(cè)趯ふ疫^(guò)程中并沒(méi)有嘗試尋找所有的參數(shù)組合传黄，所以我們不能確定這個(gè)最小值為全局最小值。

梯度下降算法的公式為：

其中：

因此队寇，我們可將表達(dá)式改寫(xiě)為：

其中α是學(xué)習(xí)率（Learning Rate）尝江，α越大則代價(jià)函數(shù)值下降得越快；反之英上，代價(jià)函數(shù)值下降得越慢炭序。表達(dá)式中 “:=” 這個(gè)符號(hào)表示賦值啤覆。

注：我們?cè)诟聟?shù)θ的值時(shí)要做到同步更新，即同時(shí)更新θ₀惭聂，θ₁的值窗声。

補(bǔ)充筆記

Gradient Descent

So we have our hypothesis function and we have a way of measuring how well it fits into the data. Now we need to estimate the parameters in the hypothesis function. That's where gradient descent comes in.

Imagine that we graph our hypothesis function based on its fields θ₀ and θ₁ (actually we are graphing the cost function of the parameter estimates). We are not graphing x and y itself, but the parameter range of our hypothesis function and the cost resulting from selecting a particular set of parameters.

We put θ₀ on the x axis and θ₁ on the y axis, with the cost function on the vertical z axis. The points on our graph will be the result of the cost function using our hypothesis with those specific theta parameters. The graph below depicts such a setup.

We will know that we have succeeded when our cost function is at the very bottom of the pits in our graph, i.e. when its value is the minimum. The red arrows show the minimum points in the graph.

The way we do this is by taking the derivative (the tangential line to a function) of our cost function. The slope of the tangent is the derivative at that point and it will give us a direction to move towards. We make steps down the cost function in the direction with the steepest descent. The size of each step is determined by the parameter α, which is called the learning rate.

The gradient descent algorithm is:

repeat until convergence:

where
j = 0, 1 represents the feature index number.

At each iteration j, one should simultaneously update the parameters θ₁, θ₂, ... , θ_n. Updating a specific parameter prior to calculating another one on the j^(th) iteration would yield to a wrong implementation.

Gradient Descent Intuition

We explored the scenario where we used one parameter θ₁ and its cost function to implement a gradient. Our formula for a single parameter was:

repeat until convergence:

On a side note, we should adjust our parameter α to ensure that the gradient descent algorithm converges in a reasonable time. Failure to converge or too much time to obtain the minimum value imply that our step size is wrong.

How does gradient descent converge with a fixed step size α?

Gradient Descent For Linear Regression

When specifically applied to the case of linear regression, a new form of the gradient descent equation can be derived. We can substitute our actual cost function and our actual hypothesis function and modify equation to:

where m is the size of the training set θ₀ a constant that will be changing simultaneously with θ₁ and x_i, y_i are values of the given training set (data).

The point of all this is that if we start with a guess for our hypothesis and then repeatedly apply these gradient descent equations, our hypothesis will become more and more accurate.

So, this is simply gradient descent on the original cost function J. This method looks at every example in the entire training set on every step, and is called batch gradient descent. Not that, while gradient descent can be susceptible to local minimum in general, the optimization problem we have posed here for linear regression has only one global, and no other local, optima; thus gradient descent always converges (assuming the learning rate α is not too large) to the global minimum. Indeed, J is a convex quadratic function. Here is an example of gradient descent as it is run to minimize a quadratic function.

The ellipses shown above are the contours of a quadratic function. Also shown is the trajectory taken by gradient descent, which was initialized at (48, 30). The x's in the figure (joined by straight lines) mark the successive values of θ that gradient descent went through as it converged to its minimum.

注：國(guó)外與國(guó)內(nèi)關(guān)于凹凸函數(shù)的定義是反的。