在昨天人工神經(jīng)網(wǎng)絡(luò)課程之后，有一位同學(xué)課下問(wèn)了一個(gè)問(wèn)題，她這學(xué)期也在學(xué)習(xí)“機(jī)器學(xué)習(xí)”課程，感覺(jué)“人工神經(jīng)網(wǎng)絡(luò)”課程的內(nèi)容與機(jī)器學(xué)習(xí)課程的內(nèi)容大同小異。究竟這些課程之間有何區(qū)別呢？弄不清楚這些自己這學(xué)期的課程很是擔(dān)心。

之所以產(chǎn)生這樣的疑問(wèn)，原因來(lái)自于這兩門(mén)課程之間的相似之處，而且隨著學(xué)科的發(fā)展它們重合度也在增加。但它們之間的差異在哪兒呢？

除了它們各自發(fā)展的理論和技術(shù)歷史和路徑不同、未來(lái)研究熱點(diǎn)和實(shí)現(xiàn)途徑差異之外，也許認(rèn)清它們之間的聯(lián)系更重要。

DJ Patil在他的一個(gè)短片中?What’s the difference between ML and NN??總結(jié)了機(jī)器學(xué)習(xí)和人工神經(jīng)網(wǎng)絡(luò)幾點(diǎn)關(guān)系：

1. 一種對(duì)人工神經(jīng)網(wǎng)絡(luò)，機(jī)器學(xué)習(xí)，人工智能之間關(guān)系的最基本看法是：人工神經(jīng)網(wǎng)絡(luò)是眾多問(wèn)題解決方案中的一種；

   
   

2. 現(xiàn)今階段你所能看到的人工神經(jīng)網(wǎng)絡(luò)大部分是一種使用大量數(shù)據(jù)訓(xùn)練的多層深度學(xué)習(xí)網(wǎng)絡(luò)，并在傳統(tǒng)的誤差反向傳播（BP）技術(shù)之上衍生出很多其他特性；

3. 對(duì)于神經(jīng)網(wǎng)絡(luò)算法的提高也使得它與機(jī)器學(xué)習(xí)方法有了很多共同之處：比如監(jiān)督學(xué)習(xí)、非監(jiān)督學(xué)習(xí)、Logistic回歸、隨機(jī)森林等。這些方法的共同之處都是通過(guò)一些訓(xùn)練數(shù)據(jù)及來(lái)尋找到一些滿足某些約束條件的函數(shù)映射。

近日，一篇來(lái)自于斯坦福大學(xué)的人工智能定義短文?Artificial Intelligence Definitions?從某一角度較為詳細(xì)的把智能相關(guān)的概念進(jìn)行了梳理，閱讀它也許可以幫你盡可能理清這個(gè)領(lǐng)域中的眾多學(xué)科之間的關(guān)系。

智能?可以被定義為在不確定、時(shí)刻變化的環(huán)境中通過(guò)學(xué)習(xí)和實(shí)施合適的技術(shù)來(lái)解決碰到的問(wèn)題或達(dá)到既定目標(biāo)的能力。而那種安全靠編程來(lái)靈活、精確、可靠工作的工廠中的機(jī)器人則不具有智能。

Intelligence?might be defined as the ability to learn and perform suitable techniques to solve problems and achieve goals, appropriate to the context in an uncertain,ever-varying world. A fully pre-programmed factory robotis flexible, accurate, and consistent but not intelligent.

人工智能?這一詞語(yǔ)是由斯坦福大學(xué)退休名譽(yù)教授 McCarthy 在1955年提出，是指：“制造出智能設(shè)備的科學(xué)和工程技術(shù)?！?多數(shù)研究是通過(guò)計(jì)算機(jī)編程使得機(jī)器表現(xiàn)出聰明，比如下象棋。但今天我們更強(qiáng)調(diào)機(jī)器能夠像人類一樣進(jìn)行學(xué)習(xí)。

Artificial Intelligence (AI), a term coined by emeritus Stanford Professor John McCarthy in 1955,was defined by him as “the science and engineering of making intelligent machines”. Much research has humans program machines to behave in a clever way, like playing chess, but, today, we emphasize machines that can learn,at least somewhat like human beings do.

自主系統(tǒng)?無(wú)需借助底層（微?。┕芾肀憧梢宰灾鲗?duì)完成特定目標(biāo)的步驟進(jìn)行規(guī)劃和決策。比如在醫(yī)院里遞送藥品的機(jī)器人可以成功穿越擁擠的走廊完成這個(gè)任務(wù)。在人工智能中的自主，與政治和生物領(lǐng)域中的自主概念并不相同。

Autonomous systems?can independently plan and decide sequences of steps to achieve a specified goal without micro-management. A hospital delivery robot must autonomously navigate busy corridors to succeed in its task. In AI, autonomy doesn’t have the sense of being self-governing common in politics or biology.

機(jī)器學(xué)習(xí)?是人工智能中研究如何通過(guò)計(jì)算機(jī)軟件，在已有的經(jīng)驗(yàn)和數(shù)據(jù)的基礎(chǔ)上提高感知、知識(shí)、思考和行動(dòng)能力。為實(shí)現(xiàn)這個(gè)目的，機(jī)器學(xué)習(xí)應(yīng)用了計(jì)算機(jī)科學(xué)、統(tǒng)計(jì)學(xué)、心理學(xué)、神經(jīng)科學(xué)、經(jīng)濟(jì)學(xué)(?) 以及控制理論。

Machine Learning (ML)?is the part of AI studying how computer agents can improve their perception,knowledge, thinking, or actions based on experience or data. For this, ML draws from computer science, statistics,psychology, neuroscience, economics and control theory.

監(jiān)督學(xué)習(xí)?計(jì)算機(jī)軟件從人類給定數(shù)據(jù)的標(biāo)簽中來(lái)進(jìn)行預(yù)測(cè)，比如從狗的圖片來(lái)預(yù)測(cè)狗的品種。?無(wú)監(jiān)督學(xué)習(xí)?則無(wú)需數(shù)據(jù)標(biāo)簽，有時(shí)是自行完成預(yù)測(cè)任務(wù)，例如預(yù)測(cè)一個(gè)句子中詞語(yǔ)的前后順序。?增強(qiáng)學(xué)習(xí)?是根據(jù)總的獎(jiǎng)賞條件來(lái)自行確定要執(zhí)行任務(wù)，比如在游戲中，無(wú)需給定哪種好的技術(shù)，它便可以自行學(xué)習(xí)。

In?supervised learning, a computer learns to predict human-given labels, such as dog breed based on labeled dog pictures;?unsupervised learning?doesnot require labels, sometimes making its own prediction tasks such as trying to predict each successive word in a sentence;?reinforcement learning?lets an agent learn action sequences that optimize its total rewards,such as winning games, without explicit examples of good techniques, enabling autonomy.

深度學(xué)習(xí)?則通過(guò)使用大型多層人工神經(jīng)網(wǎng)絡(luò)?形成類似于人腦中神經(jīng)元的層次結(jié)構(gòu)，計(jì)算它們之間連續(xù)變化的權(quán)值。這是當(dāng)今機(jī)器學(xué)習(xí)各種方法中應(yīng)用最為成功的方法。無(wú)論數(shù)據(jù)集合大小還是計(jì)算能力消耗量，它都能夠很好的推廣。

Deep Learning?is the use of large multi-layer(artificial) neural networks?that compute with continuous (real number) representations, a little like the hierarchically organized neurons in human brains. It is currently the most successful ML approach, usable for all types of ML, with better generalization from small data and better scaling to big data and compute budgets.

普通的算法需要像編寫(xiě)的計(jì)算機(jī)程序那樣有精確的執(zhí)行步驟。而人工智能算法則只有少量的用于描述學(xué)習(xí)和獎(jiǎng)勵(lì)的計(jì)算方法，大部分算法的性能則是通過(guò)對(duì)數(shù)據(jù)和經(jīng)驗(yàn)的學(xué)習(xí)來(lái)獲得。對(duì)于這種巨大的變化，b畢業(yè)于斯坦福大學(xué)的Andrej Kapathy稱為它?軟件的2.0版本。

An?algorithm?lists the precise steps to take, such as a person writes in a computer program. AI systems contain algorithms, but often just for a few parts like a learning or reward calculation method. Much of their behavior emerges via learning from data or experience,a sea change in system design that Stanford alumnus Andrej Karpathy dubbed?Software 2.0.

狹義的人工智能?是為了完成特定任務(wù)，比如語(yǔ)音、人臉識(shí)別。類人智能，或者廣義智能?則是探索更加一般性的智能，適用不同語(yǔ)境的機(jī)器。比如一些社交聊天機(jī)器人或者人與機(jī)器人的交互等。

Narrow AI?is intelligent systems for one particular thing, e.g.,?speech?or?facial recognition.Human-level AI,?or?Artificial General Intelligence (AGI), seeks broadly intelligent,context-aware machines. It is needed for effective?social chatbots?or?human-robot interaction.

聚焦人類人工智能?則是為增加人的能力、解決社會(huì)需求、從人類行為獲得啟發(fā)的人工智能，研究制作人類更加有效伙伴和工具，比如對(duì)老年人的輔助和關(guān)愛(ài)機(jī)器人。

Human-Centered Artificial Intelligence?is AI that seeks to augment the abilities of, address the societal needs of, and draw inspiration from human beings. It researches and builds effective partners and tools for people, such as a robot helper and companion for the elderly.

2020年最后一場(chǎng)智能車競(jìng)賽

在11月7~9日，安徽賽區(qū)第十五屆智能車競(jìng)賽將在合肥職業(yè)技術(shù)學(xué)院匯心湖校區(qū)舉。各個(gè)組別參賽隊(duì)伍統(tǒng)計(jì)如下：

2020年第一場(chǎng)智能車省賽

2020年第九屆四川省智能車競(jìng)賽剛剛在樂(lè)山師范學(xué)院結(jié)束。下面讓我們看看他們比賽現(xiàn)場(chǎng)情況吧。