機(jī)器學(xué)習(xí) 導(dǎo)論

什么是機(jī)器學(xué)習(xí)？ (What is Machine Learning?)

Machine learning can be vaguely defined as a computers ability to learn without being explicitly programmed, this, however, is an older definition of machine learning. A more modern definition was given by Tom Mitchell, "A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E."

可以將機(jī)器學(xué)習(xí)模糊地定義為無(wú)需明確編程即可學(xué)習(xí)的計(jì)算機(jī)能力，但是，這是機(jī)器學(xué)習(xí)的較早定義。 湯姆·米切爾(Tom Mitchell)給出了一個(gè)更現(xiàn)代的定義： “如果某計(jì)算機(jī)程序在T中的任務(wù)上的性能(由P來(lái)衡量)隨著經(jīng)驗(yàn)的提高而提高，則該計(jì)算機(jī)程序可以從經(jīng)驗(yàn)E中學(xué)習(xí)一些任務(wù)T和性能指標(biāo)P E。”

For instance, let's assume we have an algorithm that watches emails a user marks as spam and based on that observation it learns to filter out unwanted spam messages. The experience E in the above situation would be to Watch and recognize what type of mail is marked as spam. The task T would be to filter mail as spam based on the experience E. The Performance P would is the efficiency at which the algorithm filters spam mail and it would simply improve with the experience E.

例如，假設(shè)我們有一個(gè)算法可以監(jiān)視用戶標(biāo)記為垃圾郵件的電子郵件，并根據(jù)該觀察結(jié)果學(xué)會(huì)過(guò)濾掉不需要的垃圾郵件。 在上述情況下的體驗(yàn)E是觀察并識(shí)別哪種類型的郵件被標(biāo)記為垃圾郵件。 任務(wù)T是根據(jù)經(jīng)驗(yàn)E將郵件過(guò)濾為垃圾郵件。性能P將是算法過(guò)濾垃圾郵件的效率，并且會(huì)隨經(jīng)驗(yàn)E的提高而提高。

Machine learning is often confused with Artificial intelligence. Artificial intelligence is measured as the ability of a machine to behave as a human being whereas Machine learning is a subset of artificial intelligence that deals with training a machine or computer to learn from large amounts of data supplied to it.

機(jī)器學(xué)習(xí)通常與人工智能相混淆。 人工智能被衡量為機(jī)器表現(xiàn)為人類的能力，而機(jī)器學(xué)習(xí)是人工智能的子集，其處理訓(xùn)練機(jī)器或計(jì)算機(jī)以從提供給它的大量數(shù)據(jù)中學(xué)習(xí)。

Machine learning is implemented in two ways, Supervised and Unsupervised learning.

機(jī)器學(xué)習(xí)有兩種實(shí)現(xiàn)方式，監(jiān)督學(xué)習(xí)和無(wú)監(jiān)督學(xué)習(xí)。

Supervised learning is when the machine is given a specific data set along with the correct output. Here the machine is given an idea of what the output must look like with respect to the given input. Supervised learning is further classified into two subsets namely, Regression learning problems and Classification learning problems.

監(jiān)督學(xué)習(xí)是指為機(jī)器提供特定的數(shù)據(jù)集以及正確的輸出。 在這里，機(jī)器將獲得關(guān)于給定輸入的輸出外觀的概念。 監(jiān)督學(xué)習(xí)被進(jìn)一步分為兩個(gè)子集，即回歸學(xué)習(xí)問(wèn)題和分類學(xué)習(xí)問(wèn)題。

In a regression learning problem, we try and obtain predictions as a continuous function of the given input and not as a discrete value whereas in Classification learning problems we try to obtain a discrete value of the output based on previously analyzed data and the given input.

在回歸學(xué)習(xí)問(wèn)題中，我們嘗試獲取作為給定輸入的連續(xù)函數(shù)而不是離散值的預(yù)測(cè)，而在分類學(xué)習(xí)問(wèn)題中，我們嘗試基于先前分析的數(shù)據(jù)和給定輸入來(lái)獲取輸出的離散值。

In classification learning problems, on the other hand, we approach problems without any knowledge about the correct output. The required relationship between the given data and solution can be acquired by clustering the given data based on the relationship of the individual variables present in the given data.

另一方面，在分類學(xué)習(xí)問(wèn)題中，我們?cè)跊](méi)有任何正確輸出知識(shí)的情況下處理問(wèn)題。 可以通過(guò)基于給定數(shù)據(jù)中存在的各個(gè)變量的關(guān)系對(duì)給定數(shù)據(jù)進(jìn)行聚類來(lái)獲取給定數(shù)據(jù)與解決方案之間的所需關(guān)系。

Machine learning is used and implemented in various fields of application. Most of us use machine learning algorithms unknowingly in our daily lives. Some of the common applications of machine learning are, Social media services such as personalized social media and news feeds by the content is being searched for, advertisement targetting and product recommendations by monitoring products or services viewed online, email and malware filtering by monitoring the content marked as spam and content classified as malware by users, Refining search engine results to improve search result by monitoring the time spent visiting and viewing web results, personalizing home and voice assistants by monitoring users internet and web activity. Machine learning is an important aspect to predicting highly accurate solutions to problems in various fields of applications such as science, medicine and commerce and can be employed to simplify and improve the quality and rate at which problems are solved.

機(jī)器學(xué)習(xí)在各種應(yīng)用領(lǐng)域中得到使用和實(shí)現(xiàn)。 我們大多數(shù)人在日常生活中不知不覺中使用了機(jī)器學(xué)習(xí)算法。 機(jī)器學(xué)習(xí)的一些常見應(yīng)用包括：社交媒體服務(wù)(例如按內(nèi)容搜索個(gè)性化社交媒體和新聞提要)，通過(guò)監(jiān)視在線觀看的產(chǎn)品或服務(wù)來(lái)確定廣告目標(biāo)和產(chǎn)品推薦，通過(guò)監(jiān)視內(nèi)容來(lái)進(jìn)行電子郵件和惡意軟件過(guò)濾被用戶標(biāo)記為垃圾郵件和被用戶歸類為惡意軟件的內(nèi)容，通過(guò)監(jiān)視訪問(wèn)和查看Web結(jié)果所花費(fèi)的時(shí)間，通過(guò)監(jiān)視用戶的Internet和Web活動(dòng)來(lái)個(gè)性化家庭和語(yǔ)音助手來(lái)完善搜索引擎結(jié)果以改善搜索結(jié)果。 機(jī)器學(xué)習(xí)是預(yù)測(cè)諸如科學(xué)，醫(yī)學(xué)和商業(yè)等各種應(yīng)用領(lǐng)域中的問(wèn)題的高精度解決方案的重要方面，并且可以用來(lái)簡(jiǎn)化和提高解決問(wèn)題的質(zhì)量和速度。

翻譯自: https://www.includehelp.com/ml-ai/introduction-to-machine-learning.aspx

機(jī)器學(xué)習(xí) 導(dǎo)論

總結(jié)

以上是生活随笔為你收集整理的机器学习 导论_机器学习导论的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問(wèn)題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。