463
个编辑
更改
→Limitations
== Limitations ==
== 局限性Limitations ==
Although machine learning has been transformative in some fields, machine-learning programs often fail to deliver expected results.<ref>{{Cite news|url=https://www.bloomberg.com/news/articles/2016-11-10/why-machine-learning-models-often-fail-to-learn-quicktake-q-a|title=Why Machine Learning Models Often Fail to Learn: QuickTake Q&A|date=2016-11-10|work=Bloomberg.com|access-date=2017-04-10|archive-url=https://web.archive.org/web/20170320225010/https://www.bloomberg.com/news/articles/2016-11-10/why-machine-learning-models-often-fail-to-learn-quicktake-q-a|archive-date=2017-03-20}}</ref><ref>{{Cite news|url=https://hbr.org/2017/04/the-first-wave-of-corporate-ai-is-doomed-to-fail|title=The First Wave of Corporate AI Is Doomed to Fail|date=2017-04-18|work=Harvard Business Review|access-date=2018-08-20}}</ref><ref>{{Cite news|url=https://venturebeat.com/2016/09/17/why-the-a-i-euphoria-is-doomed-to-fail/|title=Why the A.I. euphoria is doomed to fail|date=2016-09-18|work=VentureBeat|access-date=2018-08-20|language=en-US}}</ref> Reasons for this are numerous: lack of (suitable) data, lack of access to the data, data bias, privacy problems, badly chosen tasks and algorithms, wrong tools and people, lack of resources, and evaluation problems.<ref>{{Cite web|url=https://www.kdnuggets.com/2018/07/why-machine-learning-project-fail.html|title=9 Reasons why your machine learning project will fail|website=www.kdnuggets.com|language=en-US|access-date=2018-08-20}}</ref>
Although machine learning has been transformative in some fields, machine-learning programs often fail to deliver expected results.<ref>{{Cite news|url=https://www.bloomberg.com/news/articles/2016-11-10/why-machine-learning-models-often-fail-to-learn-quicktake-q-a|title=Why Machine Learning Models Often Fail to Learn: QuickTake Q&A|date=2016-11-10|work=Bloomberg.com|access-date=2017-04-10|archive-url=https://web.archive.org/web/20170320225010/https://www.bloomberg.com/news/articles/2016-11-10/why-machine-learning-models-often-fail-to-learn-quicktake-q-a|archive-date=2017-03-20}}</ref><ref>{{Cite news|url=https://hbr.org/2017/04/the-first-wave-of-corporate-ai-is-doomed-to-fail|title=The First Wave of Corporate AI Is Doomed to Fail|date=2017-04-18|work=Harvard Business Review|access-date=2018-08-20}}</ref><ref>{{Cite news|url=https://venturebeat.com/2016/09/17/why-the-a-i-euphoria-is-doomed-to-fail/|title=Why the A.I. euphoria is doomed to fail|date=2016-09-18|work=VentureBeat|access-date=2018-08-20|language=en-US}}</ref> Reasons for this are numerous: lack of (suitable) data, lack of access to the data, data bias, privacy problems, badly chosen tasks and algorithms, wrong tools and people, lack of resources, and evaluation problems.<ref>{{Cite web|url=https://www.kdnuggets.com/2018/07/why-machine-learning-project-fail.html|title=9 Reasons why your machine learning project will fail|website=www.kdnuggets.com|language=en-US|access-date=2018-08-20}}</ref>
Although machine learning has been transformative in some fields, machine-learning programs often fail to deliver expected results. Reasons for this are numerous: lack of (suitable) data, lack of access to the data, data bias, privacy problems, badly chosen tasks and algorithms, wrong tools and people, lack of resources, and evaluation problems.
Although machine learning has been transformative in some fields, machine-learning programs often fail to deliver expected results. Reasons for this are numerous: lack of (suitable) data, lack of access to the data, data bias, privacy problems, badly chosen tasks and algorithms, wrong tools and people, lack of resources, and evaluation problems.
尽管机器学习在某些领域具有革命性的作用，但机器学习程序往往无法完全实现预期的效果。造成这种情况的原因有很多: 缺乏（合适的）数据、缺乏对数据的访问、存在数据偏见、隐私问题、选择不当的任务和算法、错误的工具和人员、缺乏资源以及算法评估不当等问题。
In 2018, a self-driving car from Uber failed to detect a pedestrian, who was killed after a collision. Attempts to use machine learning in healthcare with the IBM Watson system failed to deliver even after years of time and billions of investment.
In 2018, a self-driving car from Uber failed to detect a pedestrian, who was killed after a collision. Attempts to use machine learning in healthcare with the IBM Watson system failed to deliver even after years of time and billions of investment.
2018年，优步的一名自动驾驶汽车司机由于未能检测到一名行人，导致该行人在该次碰撞事故中丧生。在医疗保健中使用 IBM Watson 系统的机器学习进行尝试，但经过多年的时间和数十亿的投资也尚未能实现。
===Bias===
=== 偏差 Bias===
{{main|Algorithmic bias}}
{{main|Algorithmic bias}}
Machine learning approaches in particular can suffer from different data biases. A machine learning system trained on current customers only may not be able to predict the needs of new customer groups that are not represented in the training data. When trained on man-made data, machine learning is likely to pick up the same constitutional and unconscious biases already present in society. Language models learned from data have been shown to contain human-like biases. Machine learning systems used for criminal risk assessment have been found to be biased against black people. In 2015, Google photos would often tag black people as gorillas, and in 2018 this still was not well resolved, but Google reportedly was still using the workaround to remove all gorillas from the training data, and thus was not able to recognize real gorillas at all. Similar issues with recognizing non-white people have been found in many other systems. In 2016, Microsoft tested a chatbot that learned from Twitter, and it quickly picked up racist and sexist language. Because of such challenges, the effective use of machine learning may take longer to be adopted in other domains. Concern for fairness in machine learning, that is, reducing bias in machine learning and propelling its use for human good is increasingly expressed by artificial intelligence scientists, including Fei-Fei Li, who reminds engineers that "There’s nothing artificial about AI...It’s inspired by people, it’s created by people, and—most importantly—it impacts people. It is a powerful tool we are only just beginning to understand, and that is a profound responsibility.”
Machine learning approaches in particular can suffer from different data biases. A machine learning system trained on current customers only may not be able to predict the needs of new customer groups that are not represented in the training data. When trained on man-made data, machine learning is likely to pick up the same constitutional and unconscious biases already present in society. Language models learned from data have been shown to contain human-like biases. Machine learning systems used for criminal risk assessment have been found to be biased against black people. In 2015, Google photos would often tag black people as gorillas, and in 2018 this still was not well resolved, but Google reportedly was still using the workaround to remove all gorillas from the training data, and thus was not able to recognize real gorillas at all. Similar issues with recognizing non-white people have been found in many other systems. In 2016, Microsoft tested a chatbot that learned from Twitter, and it quickly picked up racist and sexist language. Because of such challenges, the effective use of machine learning may take longer to be adopted in other domains. Concern for fairness in machine learning, that is, reducing bias in machine learning and propelling its use for human good is increasingly expressed by artificial intelligence scientists, including Fei-Fei Li, who reminds engineers that "There’s nothing artificial about AI...It’s inspired by people, it’s created by people, and—most importantly—it impacts people. It is a powerful tool we are only just beginning to understand, and that is a profound responsibility.”
特别是机器学习方法可能会受到不同数据偏差的影响。仅针对当前客户的机器学习系统可能无法预测培训数据中没有表示的新客户群体的需求。当机器学习接受人造数据的训练时，很可能会挑选出社会中已经存在的同样的宪法和无意识的偏见。从数据中学到的语言模型已经被证明包含了类似人类的偏见。用于犯罪风险评估的机器学习系统被发现对黑人有偏见。在2015年，谷歌照片经常把黑人标记为大猩猩，而在2018年，这个问题仍然没有得到很好的解决，但据报道，谷歌仍然在使用变通方法从训练数据中删除所有大猩猩，因此根本无法识别真正的大猩猩。在许多其他系统中也发现了识别非白人的类似问题。2016年，微软测试了一个从 Twitter 上学来的聊天机器人，它很快就学会了种族主义和性别歧视的语言。由于这些挑战，机器学习的有效应用可能需要更长的时间才能被其他领域采用。人工智能科学家越来越关注机器学习中的公平问题，即减少机器学习中的偏见，推动机器学习为人类的利益服务。李等人提醒工程师，“人工智能没有任何人为的东西... ... 它受到人的启发，由人创造，最重要的是，它会影响人。”。这是一个强大的工具，我们才刚刚开始理解，这是一项意义深远的责任。”
特别是机器学习方法可能会受到不同数据偏差的影响。仅针对当前客户的机器学习系统可能无法预测培训数据中没有表示的新客户群体的需求。当机器学习接受人造数据的训练时，很可能会挑选出社会中已经存在的同样的宪法和无意识的偏见。从数据中学到的语言模型已经被证明包含了类似人类的偏见。用于犯罪风险评估的机器学习系统被发现对黑人有偏见。在2015年，谷歌照片经常把黑人标记为大猩猩，而在2018年，这个问题仍然没有得到很好的解决，但据报道，谷歌仍然在使用变通方法从训练数据中删除所有大猩猩，因此根本无法识别真正的大猩猩。在许多其他系统中也发现了识别非白人的类似问题。2016年，微软测试了一个从 Twitter 上学来的聊天机器人，它很快就学会了种族主义和性别歧视的语言。由于这些挑战，机器学习的有效应用可能需要更长的时间才能在其他领域内广泛应用。人工智能科学家越来越关注机器学习中的公平问题，即减少机器学习中的偏见，推动机器学习为人类的利益服务。李等人提醒工程师们：“人工智能没有任何人为的东西... ... 它受到人的启发，由人创造，最重要的是，它会影响人。”，”这是一个强大的工具，我们才刚刚开始理解，这是一项意义深远的责任。”
== Model assessments ==
== Model assessments ==