案例学习：预测房价

Loading...

来自 华盛顿大学 的课程

机器学习：回归

3443 评分

案例学习：预测房价

从本节课中

Simple Linear Regression

Our course starts from the most basic regression model: Just fitting a line to data. This simple model for forming predictions from a single, univariate feature of the data is appropriately called "simple linear regression".<p> In this module, we describe the high-level regression task and then specialize these concepts to the simple linear regression case. You will learn how to formulate a simple regression model and fit the model to data using both a closed-form solution as well as an iterative optimization algorithm called gradient descent. Based on this fitted function, you will interpret the estimated model parameters and form predictions. You will also analyze the sensitivity of your fit to outlying observations.<p> You will examine all of these concepts in the context of a case study of predicting house prices from the square feet of the house.

- Emily FoxAmazon Professor of Machine Learning

Statistics - Carlos GuestrinAmazon Professor of Machine Learning

Computer Science and Engineering

Okay, so the first regression task that we have is

we have to figure out what model are we gonna use.

Are we gonna assume that there is just a constant relationship between square

footage and price?

That means regardless of the size of the room,

we are expecting every house to sell for the same amount.

Well that's probably not a great model.

Are we gonna assume that there's some linear relationship?

So as I increase square footage,

my price increases at the same rate as I'm increasing square footage.

Or I'm I gonna assume that there's some quadratic fit or

some higher order polynomial fit, or the list of models I could consider

is very long and that's what this course is partially gonna be about.

Exploring different options that we have for models of our data.

Okay, so one task is out of the space of all these models that we might consider.

Which is the one that we should use for a given dataset and task that we have?

Okay, but now, let's assume that we

have selected the model we're gonna use,

in this case, here, we're assuming that

we're gonna use just a quadratic fit, so

assume Model, f of x, is a quadratic function.

Then our next task is gonna be to estimate a specific quadratic fit to the data.

Okay, so a model just specifies the form of something,

it's gonna be defined in terms of some set of parameters.

And then we're gonna have to estimate what the specific fit is from the data.

So for example, here,

this is our estimated quadratic fit.

And we'll call it f hat of x, this is our estimated

function that's fit from our specific dataset.

Or this is another function that we could have fit.

And we'll talk about the way in which we're gonna fit

functions to data in this course.

Okay, but the point is that first we have to choose a model, then we have to

provide some procedure, some algorithm, for fitting that model to the data.

And coming up with a specific curve that we're gonna use for

our tasks such as prediction.

[MUSIC]