谷歌人工智能文章中英文對(duì)照.doc

1 5 非專業(yè)人士翻譯 如有錯(cuò)誤請(qǐng)諒解 Google s AI Reasons Its Way around the London Underground 谷歌人工智能推導(dǎo)出環(huán)繞 倫敦地鐵系統(tǒng)的路線 DeepMind s latest technique uses external memory to solve tasks that require logic and reasoning a step toward more humanlike AI 深度思維最新技術(shù)使用了外部存儲(chǔ)來(lái)解決需要邏輯思 維和推理能力的任務(wù) By Elizabeth Gibney Nature magazine on October 14 2016 伊麗莎白 吉布尼 2016 年 10 月 14 日發(fā)表于 自然 雜志 Artificial intelligence AI systems known as neural networks can recognize images translate languages and even master the ancient game of Go But their limited ability to represent complex relationships between data or variables has prevented them from conquering tasks that require logic and reasoning 人工智能 AI 系統(tǒng)被認(rèn)為是神經(jīng)網(wǎng)絡(luò) 可以識(shí)別圖片 翻譯 甚 至精通古老的游戲 但他們描繪數(shù)據(jù)或變量之間的復(fù)雜關(guān)系的能力有 限 這妨礙了他們克服需要邏輯思維和推理能力的任務(wù) In a paper published in Nature on October 12 the Google owned company DeepMind in London reveals that it has taken a step towards overcoming this hurdle by creating a neural network with an external memory The combination allows the neural network not 2 5 only to learn but to use memory to store and recall facts to make inferences like a conventional algorithm This in turn enables it to tackle problems such as navigating the London Underground without any prior knowledge and solving logic puzzles Though solving these problems would not be impressive for an algorithm programmed to do so the hybrid system manages to accomplish this without any predefined rules 在 10 月 12 日 自然 雜志中發(fā)表的一篇論文中 谷歌在倫敦的 子公司深度思維展示了他們通過結(jié)合外部存儲(chǔ)創(chuàng)造了一個(gè)神經(jīng)網(wǎng)絡(luò) 來(lái)進(jìn)一步克服這些障礙 這種和外部存儲(chǔ)的結(jié)合不僅允許神經(jīng)網(wǎng)絡(luò)學(xué) 習(xí) 還可以通過存儲(chǔ)器來(lái)存儲(chǔ)和回憶事件 并以此來(lái)像正常情況那樣 做推斷 這反過來(lái)能夠讓它解決難題 比如在沒有任何經(jīng)驗(yàn)的情況下 操控倫敦地鐵 比如解決邏輯謎題 盡管對(duì)于一個(gè)算法程序來(lái)說(shuō)做到 這點(diǎn)并不會(huì)令人印象深刻 但這個(gè)混合系統(tǒng)在沒有任何先決條件的情 況下做到了這點(diǎn) Although the approach is not entirely new DeepMind itself reported attempting a similar feat in a preprint in 2014 the progress made in this paper is remarkable says Yoshua Bengio a computer scientist at the University of Montreal in Canada 雖然這個(gè)方法不是一個(gè)全新的技術(shù) 深度思維自己就在 2014 年報(bào)告過他們嘗試了一種相似的技術(shù) 但 在論文中的這個(gè)進(jìn)步是 非凡的 加拿大蒙特利爾的計(jì)算機(jī)學(xué)家本吉奧 本希奧贊嘆道 MEMORY MAGIC 記記憶憶 魔魔法法 A neural network learns by strengthening connections between virtual neuron like units Without a memory such a network might 3 5 need to see a specific London Undeground map thousands of times to learn the best way to navigate the tube 神經(jīng)網(wǎng)絡(luò)通過加強(qiáng)虛擬神經(jīng)元之間的聯(lián)系來(lái)學(xué)習(xí) 如果沒有存儲(chǔ) 器 這樣一個(gè)網(wǎng)絡(luò)可能需要看一副特定的倫敦地鐵地圖數(shù)千次來(lái)學(xué)習(xí) 最佳路線 DeepMind s new system which they call a differentiable neural computer can make sense of a map it has never seen before It first trains its neural network on randomly generated map like structures which could represent stations connected by lines or other relationships in the process learning how to store descriptions of these relationships in its external memory as well as answer questions about them Confronted with a new map the DeepMind system can write these new relationships connections between Underground stations in one example from the paper to memory and recall it to plan a route 深度思維的新系統(tǒng) 他們稱它為微分神經(jīng)計(jì)算機(jī) 可以理解 它從未見過的地圖 第一次訓(xùn)練神經(jīng)網(wǎng)絡(luò)是在隨機(jī)生成的類似結(jié)構(gòu)的 地圖上 被鐵路線鏈接的車站 或者其他關(guān)系 在這個(gè)過程中學(xué)習(xí)如 何將這些關(guān)系的描述存儲(chǔ)在它的外部存儲(chǔ)器并且回答問題 面對(duì)一個(gè) 新的地圖 深度思維的系統(tǒng)可以把這些新關(guān)系 按照一個(gè)圖紙上例 子來(lái)連接各地鐵站之間的關(guān)系 寫到存儲(chǔ)器 并能夠回憶這些關(guān)系 然后計(jì)劃路線 DeepMind s AI system used the same technique to tackle puzzles that require reasoning After training on 20 different types of question and answer problems it learnt to make accurate deductions For example the system deduced correctly that a ball is in a playground having been informed that John picked up the football and John is in the playground It got such problems right more than 96 of the time The system performed better than recurrent neural networks which also have a memory but one that is in the fabric of the network itself and so is less flexible than an external memory 4 5 深度思維的人工智能系統(tǒng)使用同樣的方法來(lái)處理需要推理能力的 智力游戲 在通過 20 種不同類型的問答訓(xùn)練之后 它學(xué)會(huì)了做出準(zhǔn)確 的推論 例如 系統(tǒng)通過被告之 約翰抓著足球 和 約翰在操場(chǎng)上 準(zhǔn)確的推斷出一個(gè)球在操場(chǎng)上 答對(duì)問題的概率超過了 96 這個(gè)系統(tǒng) 的效率比擁有一個(gè)內(nèi)部存儲(chǔ)器的周期神經(jīng)網(wǎng)絡(luò)更高 也更靈活 Although the DeepMind technique has proven itself on only artificial problems it could be applied to real world tasks that involve making inferences from huge amounts of data This could solve questions whose answers are not explicitly stated in the data set says Alex Graves a computer scientist at DeepMind and a co author on the paper For example to determine whether two people lived in the same country at the same time the system might collate facts from their respective Wikipedia pages 雖然深度思維的技術(shù)已被證明只針對(duì)人工問題 但它能夠被應(yīng)用 到需要通過海量數(shù)據(jù)來(lái)進(jìn)行推斷的真實(shí)世界的工作 這能夠解決那些 在數(shù)據(jù)中沒有明確答案的問題 來(lái)自深度思維的計(jì)算機(jī)科學(xué)家 研究 報(bào)告的合著者 亞歷克斯 格雷夫斯介紹說(shuō) 例如 對(duì)于判斷兩人是 否在同一時(shí)間住在同一個(gè)國(guó)家 系統(tǒng)可能會(huì)核對(duì)他們各自在維基百科 上的事項(xiàng) Although the puzzles tackled by DeepMind s AI are simple Bengio sees the paper as a signal that neural networks are advancing beyond mere pattern recognition to human like tasks such as reasoning This extension is very important if we want to ap