當(dāng)前位置：首頁(yè) > 编程资源 > 编程问答 >内容正文

编程问答

机器学习与分布式机器学习_机器学习的歧义

發(fā)布時(shí)間：2023/12/15 编程问答 50 豆豆

生活随笔收集整理的這篇文章主要介紹了机器学习与分布式机器学习_机器学习的歧义小編覺(jué)得挺不錯(cuò)的,現(xiàn)在分享給大家,幫大家做個(gè)參考.

機(jī)器學(xué)習(xí)與分布式機(jī)器學(xué)習(xí)

超越最高精度 (Beyond Achieving Top Accuracy)

We are familiar with the idea of using machine learning to make predictions and inferences to high accuracies. This is after all a big part of what machine learning is expected to do.

我們熟悉使用機(jī)器學(xué)習(xí)對(duì)高精度進(jìn)行預(yù)測(cè)和推斷的想法。畢竟，這是機(jī)器學(xué)習(xí)所期望做的很大一部分。

Interestingly and importantly, we can leverage further on machine learning models. Beyond using the model to make top accuracy predictions, we can use it to create Uncertainty, Ambiguity or even Contention.

有趣且重要的是，我們可以進(jìn)一步利用機(jī)器學(xué)習(xí)模型。除了使用模型做出最高的準(zhǔn)確性預(yù)測(cè)外，我們還可以使用它來(lái)創(chuàng)建不確定性，歧義甚至競(jìng)爭(zhēng)。

我們不喜歡清晰和確定性嗎？ (Don’t We Like Clarity and Certainty?)

Not all the time. For good motivations,

并非一直如此。為了好的動(dòng)力

Testers may want to create examples that are uncertain, so we can stress test a system for its result or even its decision on borderline inputs.
測(cè)試人員可能想創(chuàng)建不確定的示例，因此我們可以對(duì)系統(tǒng)的結(jié)果甚至是對(duì)邊界輸入的決策進(jìn)行壓力測(cè)試。
Designers may be interested to visualise a hybrid prototype that combines existing products, but which we do not have a definite specification right now.
設(shè)計(jì)人員可能想對(duì)結(jié)合了現(xiàn)有產(chǎn)品的混合原型進(jìn)行可視化，但是我們目前尚無(wú)明確的規(guī)格。
Trainers may want to create content that is ambiguous so that there are no straightforward answers, which therefore encourage participants to engage in debate.
培訓(xùn)人員可能希望創(chuàng)建含糊不清的內(nèi)容，以便沒(méi)有簡(jiǎn)單的答案，因此鼓勵(lì)參與者進(jìn)行辯論。

The applications are plentiful.

應(yīng)用程序很多。

真實(shí)的例子 (Real Example)

Let us use the MNIST dataset that contains images of the ten digits from 0 to 9. We then train a model to at least 98% accuracy, that is, a model that correctly predicts images to their corresponding digits at least 98% of the time.

讓我們使用包含從0到9的十個(gè)數(shù)字的圖像的MNIST數(shù)據(jù)集。然后，我們訓(xùn)練一個(gè)模型至少達(dá)到98％的準(zhǔn)確度，即，一個(gè)模型至少在98％的時(shí)間內(nèi)正確地將圖像預(yù)測(cè)為其對(duì)應(yīng)的數(shù)字。

Instead of stopping here, we further leverage on the model — we use the model to apply ambiguity to the digits.

除了在這里止步不前，我們進(jìn)一步利用模型- 我們使用模型對(duì)數(shù)字應(yīng)用歧義。

Let us confuse the digit “3” with “8”.

讓我們將數(shù)字“ 3”與“ 8”混淆。

The machine learning model filled in the blanks. It dotted two spots on the left of 3. Now, it is not unreasonable for a person to contend this 3 as an 8.

機(jī)器學(xué)習(xí)模型填補(bǔ)了空白。它在3的左側(cè)點(diǎn)了兩個(gè)點(diǎn)。現(xiàn)在，一個(gè)人將3視為8并不是不合理的。

Let us confuse the digit “4” with “9”.

讓我們將數(shù)字“ 4”與“ 9”混淆。

The model attempted to build a roof over the 4 to make it closer to a 9. The model is conscious not to complete the entire roof as its goal is to make the digit uncertain between 4 and 9.

該模型試圖在4上建立屋頂以使其更接近9。該模型有意識(shí)地不完成整個(gè)屋頂，因?yàn)樗哪繕?biāo)是使數(shù)字在4到9之間不確定。

Let us confuse the digit “6” with “5”.

讓我們將數(shù)字“ 6”與“ 5”混淆。

The model cut 6 in the middle and pulled out the resulting loose end to create the hook in 5. Additionally, it sketched a short stroke at the top of 6 to make it look like 5. It is now uncertain if the digit is 6 or 5.

模型在中間切出6，然后拉出產(chǎn)生的松動(dòng)端，以在5中創(chuàng)建鉤子。此外，它還在6的頂部繪制了短筆畫(huà)以使其看起來(lái)像5。現(xiàn)在不確定數(shù)字是6還是5。 5，

模型是如何做到的？ (How does the Model Do It?)

By training the model to an accuracy of 98%, it has understood what digit images should look like. We could then ask it to engineer its knowledge and show us how digits that are uncertain, ambiguous or contentious look like.

通過(guò)將模型訓(xùn)練到98％的準(zhǔn)確度，它已經(jīng)了解了數(shù)字圖像的外觀。然后，我們可以要求它設(shè)計(jì)知識(shí)，并向我們展示不確定，模棱兩可或有爭(zhēng)議的數(shù)字。

For example, it is akin to asking “Could you create something that is in between “5” and “6” since you already know how they individually look like?”

例如，這類似于詢問(wèn)“您是否可以創(chuàng)建介于“ 5”和“ 6”之間的內(nèi)容，因?yàn)槟呀?jīng)知道它們的外觀？”

最后的想法 (Last Thoughts)

There are many good motivations for creating uncertainty, ambiguity or even contention, some of which have been explained above. Almost surely, there are also ways that uncertainty, ambiguity and contention can be used to harm or exploit! The following question probably gives us a better appreciation — could uncertain data be used as an attack against my automated system and cause it to produce an undesired decision?

產(chǎn)生不確定性，歧義甚至爭(zhēng)執(zhí)的動(dòng)機(jī)很多，上面已經(jīng)解釋了其中的一些動(dòng)機(jī)。幾乎可以肯定，不確定性，模糊性和爭(zhēng)用還可以通過(guò)其他方式來(lái)?yè)p害或利用！以下問(wèn)題可能會(huì)給我們帶來(lái)更好的理解-不確定的數(shù)據(jù)是否可以用作對(duì)我的自動(dòng)化系統(tǒng)的攻擊，并導(dǎo)致產(chǎn)生不希望的決策？

翻譯自: https://medium.com/@ryyr/machine-learning-for-ambiguity-dbb99db613af