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...

來自 University of Washington 的課程

机器学习基础：案例研究

8373 個評分

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，

從本節課中

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]

Okay, so far we've assumed that the only feature relevant to the value of my house,

is the square feet of the house.

But I dig into the data set a little bit more, and

I look at a house that's supposed to be very similar to mine.

It's a house that has very similar square footage.

And so, this house is definitely making an influence on what the predictions are for

my house.

But I look at it and I say, and I look at the specific listing, and

it shows that that house only had one bathroom.

It's actually quite a big house, only one bathroom.

And I say that's really, really not comparable to my house,

which has three bathrooms.

So that the value of that house, that house sales price

really shouldn't be indicative of what my house sales price should be.

So instead what we can think about doing is adding more features.

So instead of just looking at the relationship between square feet and

price, we can add number of bathrooms.

And now for

each one of the listings that I looked at before, I'm gonna have to go through and

record how many square feet that house had, and the number of bathrooms.

And I'm gonna plot each of these points in the 3D space.

Okay? So it's this hyper cube of square

feet versus bathrooms versus price.

And now instead of fitting a line to the data, if I'm thinking about just a very

simple model, I can think about fitting a hyper plane.

Okay, so it's just a slice through the space.

We're here.

This is the equation of the hyper plane, and this is the equation of this plane.

So we have w0, which is our intercept, just where this plane lives up and

down on the y-axis.

And we have w1 times the number of square feet, and

w2 times the number of bathrooms.

But a question is where do we stop?

Do we just want to include the number of bathrooms as our additional feature?

There are lots of things we could think about including.

We could think about in addition to our square feet, number of bathrooms,

there's the number of bedrooms, the lot size, how old the house is, and

the list goes on and on.

In terms of different properties of the house that could be influential in

assessing it's value.

But we're gonna actually hold off on this question of looking at which features

are important for this regression task, until we get to the regression course.

So go to the regression course to learn more about this topic.

[MUSIC]