大工至善|大学至真分享 http://blog.sciencenet.cn/u/lcj2212916

博文

[转载]【读书1】【2017】MATLAB与深度学习——机器学习(2)

已有 610 次阅读 2018-9-8 10:33 |系统分类:生活其它|文章来源:转载

 

深度学习近年来备受关注,因为它很好地解决了一些挑战人工智能的问题。

Deep Learning has been in the spotlightrecently as it has proficiently solved some problems that have challengedArtificial Intelligence.

 

深度学习的性能肯定在许多领域都是突出的。

Its performance surely is exceptional inmany fields.

 

然而,它也面临着局限性。

However, it faces limitations as well.

 

深度学习的局限性源于其祖先从机器学习中继承的基本概念。

The limitations of Deep Learning stems fromits fundamental concepts that has been inherited from its ancestor, MachineLearning.

 

作为一种机器学习,深度学习不能回避机器学习所面临的基本问题。

As a type of Machine Learning, DeepLearning cannot avoid the fundamental problems that Machine Learning faces.

 

这就是为什么我们需要在讨论深度学习的概念之前回顾机器学习。

That is why we need to review MachineLearning before discussing the concept of Deep Learning.

 

什么是机器学习?

What Is Machine Learning?

 

简而言之,机器学习是一种涉及数据的建模技术。

In short, Machine Learning is a modelingtechnique that involves data.

 

对于初学者来说,这个定义可能太短,无法捕捉到它的准确含义。

This definition may be too short forfirst-timers to capture what it means.

 

所以,让我详细说明一下。

So, let me elaborate on this a little bit.

 

机器学习是一种从“数据”中找出“模型”的技术。

Machine Learning is a technique thatfigures out the “model” out of “data.”

 

这里的“数据”字面意思是诸如文档、音频、图像等信息。

Here, the data literally means informationsuch as documents, audio, images, etc.

 

“模型”是机器学习的最终产物。

The “model” is the final product of MachineLearning.

 

在我们深入模型之前,让我稍微偏离一下话题。

Before we go further into the model, let medeviate a bit.

 

机器学习的定义仅仅解决了数据和模型的概念,为何与“学习”无关呢?

Isn’t it strange that the definition ofMachine Learning only addresses the concepts of data and model and has nothingto do with “learning”?

 

从术语名称本身反映了该技术用于分析数据,并发现相应的模型,而不是用人脑来实现。

The name itself reflects that the techniqueanalyzes the data and finds the model by itself rather than having a human doit.

 

我们称之为“学习”,因为这个过程类似于用数据训练来解决模型的问题。

We call it “learning” because the processresembles being trained with the data to solve the problem of finding a model.

 

因此,机器学习在建模过程中使用的数据被称为“训练”数据。

Therefore, the data that Machine Learninguses in the modeling process is called “training” data.

 

1-2描述了机器学习过程中发生的事情。

Figure 1-2 illustrates what happens in theMachine Learning process.

 


1-2 机器学习过程中究竟发生了哪些事情What happens duringthe machine learning process

 

现在,让我们继续讨论这个模型。

Now, let’s resume our discussion about themodel.

 

事实上,这个模型只不过是我们想要达到的最终产品。

Actually, the model is nothing more thanwhat we want to achieve as the final product.

 

——本文译自Phil Kim所著的《Matlab Deep Learning》



更多精彩文章请关注微信号:qrcode_for_gh_60b944f6c215_258.jpg




http://blog.sciencenet.cn/blog-69686-1133583.html

上一篇:[转载]【源码】Floyd算法——寻找网络最短路径的Matlab程序
下一篇:[转载]【源码】基于MATLAB的柑橘果实分类图像朴素贝叶斯分类器

0

该博文允许注册用户评论 请点击登录 评论 (0 个评论)

数据加载中...

Archiver|手机版|科学网 ( 京ICP备14006957 )

GMT+8, 2018-11-18 03:57

Powered by ScienceNet.cn

Copyright © 2007- 中国科学报社

返回顶部