1.a.1 机器学习的整体概念

We say that a machinelearnswith respect to a particulartaskT, performancemetric P, and type ofexperience E, if the systemreliably improves its performance P at task T, following experience E.

Machine?

No. 这里的machine是指计算机,手机等具有运算能力的智能设备。正如上文所述,机器学习和人的学习一样都是不断完善自我,提高工作效率的过程。此处“机器学习”的过程和我们写的入门级“if-else”语句的区别可用下图表达:

(注:performance:性能)

显然,“机器学习”不是死板的从头到尾然后“return 0;”结束(假设你们都用过C/C++),而是程序在执行的过程中不断的对算法自我优化,即“学习”,以达到提高性能(针对特定的task T)的目的。

上面提到的performance P, task T, experience E.经常容易被混淆,尤其是不能准确说出什么是E。下面举几个例子说明

Question 1 : Suppose we feed a learning algorithm a lot of historical weather data, and haveit learn to predict weather. In this setting, what is “E”?(注. “feed”一个算法就是让这个算法学习)

A. The weather prediction task.

B. The probability of it correctly predicting a future date's weather.

C. The process of the algorithm examining a large amount of historical weather data.

Answer:

P is The probability of it correctly predicting a future date's weather.

T is The weather prediction task.

E is The process of the algorithm examining a large amount of historical weatherdata.

Question 2: Playing Chess

A. Playing chess games

B. Percentage of games won against opponent

C. Playing practice games against itself

Answer:

T: Playing chess games

P: Percentage of games won against opponent

E: Playing practice games against itself

Question 3: Recognize Hand Written Words

A. Recognize Hand Written Digits

B. Percentage of words correctly classified

C. Database of human labeled images of hand written words

Answer:

T: Recognize Hand Written Digits

P: Percentage of words correctly classified

E: Database of human labeled images of hand written words

总结:

A. 学习

在经验的基础上提高性能

特别的,许多机器学习的学习过程是建立在数据的基础上的(数据科学的重要性之一)

B. 机器学习在计算机科学中的地位

某些算法的设计对人来说太难了,机器学习会大大减低实现算法的难度

程序需要不断适应环境,不得不通过学习来优化自身

第一次写有点技术含量的推文,有点小激动。由于本人水平有限,文章难免有不妥之处,恳请专家、同行与广大读者提出宝贵意见。

  • 发表于:
  • 原文链接https://kuaibao.qq.com/s/20180825G1EUIQ00?refer=cp_1026
  • 腾讯「云+社区」是腾讯内容开放平台帐号(企鹅号)传播渠道之一,根据《腾讯内容开放平台服务协议》转载发布内容。
  • 如有侵权,请联系 yunjia_community@tencent.com 删除。

扫码关注云+社区

领取腾讯云代金券