你是否好奇数据可以告诉你什么？你是否想在关于机器学习促进商业的核心方式上有深层次的理解？你是否想能同专家们讨论关于回归，分类，深度学习以及推荐系统的一切？在这门课上，你将会通过一系列实际案例学习来获取实践经历。在这门课结束的时候，

Loading...

From the course by University of Washington

机器学习基础：案例研究

7082 ratings

你是否好奇数据可以告诉你什么？你是否想在关于机器学习促进商业的核心方式上有深层次的理解？你是否想能同专家们讨论关于回归，分类，深度学习以及推荐系统的一切？在这门课上，你将会通过一系列实际案例学习来获取实践经历。在这门课结束的时候，

From the lesson

Regression: Predicting House Prices

This week you will build your first intelligent application that makes predictions from data.<p>We will explore this idea within the context of our first case study, predicting house prices, where you will create models that predict a continuous value (price) from input features (square footage, number of bedrooms and bathrooms,...). <p>This is just one of the many places where regression can be applied.Other applications range from predicting health outcomes in medicine, stock prices in finance, and power usage in high-performance computing, to analyzing which regulators are important for gene expression.</p>You will also examine how to analyze the performance of your predictive model and implement regression in practice using an iPython notebook.

- Carlos GuestrinAmazon Professor of Machine Learning

Computer Science and Engineering - Emily FoxAmazon Professor of Machine Learning

Statistics

[MUSIC]

So now we have some training data and some desk data.

And what we can do next is #Build,

there's a sub-header here,

#Build the regression model.

So GraphLab Create has many algorithms pre-implemented for

various machinery and tasks or you can implement your own if you like.

And during the sequence of the specialization,

you're gonna write many algorithms of your own, but for this course,

we're just gonna use pre-implemented algorithms so

can really try to understand how machine learning works.

So what you can do with machine learning?

So, we're gonna do is a build a square foot prediction model.

Just call it sqft_model.

And in GraphLab, there something called linear regression.

So, linear regression is what we learned, what we explored and

talked about, and we understood quite well during the module that Ann was teaching.

And so if you just say linear _regression.create,

it's going to create a linear regression model for you.

That creates the traits with projects GraphLab Create, and

create is also the verb to use to create various things with GraphLab Create.

So we're now going to create a linear regression model, and

what I have to give as input I give it the training data.

So here's the training data.

You have to say what variable you're trying to predict.

So that's the target.

So I'm gonna say the target is going to be price.

So I'm trying to predict price.

And the question is what are the features, what is the input.

So, I'm going to say that the features, and by the way, you can use top complete

even if [INAUDIBLE] parameters of function, which is pretty cool.

So, the features is a list of the features.

If you don't give anything, it will use all features, all columns of the data.

But here, the only feature input we want to use is square feet of living space.

And that is a string that I need to put in, so

I forgot to put it in quotes, so let me fix that real quick here.

So open quotes,

close quotes.

All right, very good.

So let me just reread that for us.

So sqft_model graphlab.linear_regression.create on

the training data to predict the target price using features sqft of living.

Let's go an execute this.

On training data training, I wrote data twice here.

You see the arrow came out here, training data and the score data.

So now let's execute.

So what's happening here?

It's done actually, but what we did was create a linear regression model,

it used a particular algorithm called Newton's Method.

But GraphLab Create chooses an algorithm automatically for

you unless you tell it which one to use.

And this algorithm allows you to make some predictions on the data.

[MUSIC]

Coursera provides universal access to the world’s best education,
partnering with top universities and organizations to offer courses online.