Random Forests

Loading...

來自 University of Washington 的課程

实用预测分析：模型与方法

272 個評分

University of Washington

实用预测分析：模型与方法

272 個評分

課程 2（共 4 門，Specialization Data Science at Scale）

Statistical experiment design and analytics are at the heart of data science. In this course you will design statistical experiments and analyze the results using modern methods. You will also explore the common pitfalls in interpreting statistical arguments, especially those associated with big data. Collectively, this course will help you internalize a core set of practical and effective machine learning methods and concepts, and apply them to solve some real world problems. Learning Goals: After completing this course, you will be able to: 1. Design effective experiments and analyze the results 2. Use resampling methods to make clear and bulletproof statistical arguments without invoking esoteric notation 3. Explain and apply a core set of classification methods of increasing complexity (rules, trees, random forests), and associated optimization methods (gradient descent and variants) 4. Explain and apply a set of unsupervised learning concepts and methods 5. Describe the common idioms of large-scale graph analytics, including structural query, traversals and recursive queries, PageRank, and community detection

從本節課中

Supervised Learning

Follow a tour through the important methods, algorithms, and techniques in machine learning. You will learn how these methods build upon each other and can be combined into practical algorithms that perform well on a variety of tasks. Learn how to evaluate machine learning methods and the pitfalls to avoid.

Bootstrap Revisited4:12

Ensembles, Bagging, Boosting4:11

Boosting Walkthrough5:07

Random Forests3:08

Random Forests: Variable Importance5:25

Summary: Trees and Forests2:48

與講師見面

Bill Howe
Director of Research
Scalable Data Analytics

[MUSIC]

So we talked about boosting and bagging and

ensemble methods and drawing bootstrap samples.

And so some of these ideas we can put together and apply them to decision trees

in order to create an ensemble method based on decision trees.

And once this ensemble method is Random Forest due to in 2001, and so

this one is really, really popular and really, really powerful.

Okay, so it's a good one to know.

So, k times your going to repeat the following procedure, okay.

Where k is a parameter that you'll specify.

You'll draw a bootstrap sample from the data set and

then train a decision tree on that bootstrap sample as follows.

So until the tree is some maximum size that is usually set pretty conservatively,

you'll choose a leaf node, iterate over all the leaf nodes.

And for each leaf node,

select m attributes at random out of the p that are available.

So you're not gonna use all, you're not gonna consider all the attributes

when you're deciding how to spit a node.

You're just gonna select m at random and consider among those.

And then pick the best attribute, or split of the attribute,

as usual using entropy or other methods and we'll talk about one more, okay?

So now you have a tree that only involves m of the attributes and

was derived from only part of the data set.

Now you measure the out-of-bag error.

So what's the out-of-bag error?

Well, when you draw a bootstrap sample remember that you might get

duplicates, right.

So you have a dataset of size n,

you're gonna draw a bootstrap sample also size n, but whenever you get a duplicate

that means that some other data item was left out of the sample, okay.

So take all those data items that were left out of the sample and

that becomes your test set, all right?

So now evaluate the error on that test set, and use this for,

you can use this for various purposes, okay.

You don't necessarily need to use it to, make decisions as part

of the algorithm but it's computed on the fly as part of the algorithm for

various other purposes, okay?

So, you can do this to estimate the strength

which is the inverse of the error rate.

You can do this to measure the correlation between random trees.

Right, so each one of these is sort of a different tree using

potentially different attributes and a different subset of the data.

But you can imagine that just by chance, you might get a bunch of trees that all do

the exact same thing and make all the same decisions, right.

This would lower the power,

lower the strength of the overall ensemble classifier.

And so you don't want that.

And so this gives you a running measure of how correlated things are, okay?

And it also gives you a measure of variable importance which we'll talk about

in the next slide.

Okay so finally at the end of this you have k trees that are hopefully more or

less independent, and

now you can classify data points coming in by just a majority vote among them.

[MUSIC]

探索我們的目錄

免費加入並獲得個性化推薦、更新和優惠。

Coursera 致力於普及全世界最好的教育，它與全球一流大學和機構合作提供在線課程。

© 2018 Coursera Inc. 保留所有權利。

通過 App Store 下載

通過 Google Play 獲取