本文共 3972 字，大约阅读时间需要 13 分钟。
机器学习与分布式机器学习
by Siraj Raval
通过Siraj Raval
Machine Learning has traditionally been a technology that only PhDs and institutions with lots of financial resources could utilize. But nowadays, there are so many tools out there that allow anyone to get started learning Machine Learning. No excuses!
传统上，机器学习是只有拥有大量财务资源的博士和机构才能利用的技术。 但是如今，有太多工具可供任何人开始学习机器学习。 没有理由！
In this blog post, I’ll highlight the four foundation stones of Machine Learning, and how each of them has been democratized in the past few years.
在这篇博客文章中，我将重点介绍机器学习的四个基础，以及它们在过去几年中如何实现民主化。
If you want to stay up to date with my latest AI content, make sure to to my YouTube channel.
如果您想了解我最新的AI内容，请确保我的YouTube频道。
The four foundation stones of Machine Learning are data, computations, algorithms, and education.
机器学习的四个基础是数据 ， 计算 ， 算法和教育 。
Check out amazing list of public datasets on GitHub. And if that’s not enough, has some amazingly polished datasets available as well. Whether you are using publicly available datasets, or scraping data from the web via Python libraries like , everyone has access to quality datasets now.
在GitHub上查看惊人的公共数据集列表。 如果这还不够的话， 还提供了一些令人惊讶的精美数据集。 无论您是使用公开可用的数据集，还是通过诸如类的Python库从Web上刮取数据，每个人现在都可以访问高质量的数据集。
Of course, the big tech companies have walled data gardens of their own, but decentralized startups like are working hard to create services to allow data scientists to train their models on that data as well.
当然，大型科技公司已经建立了自己的数据花园，但是像这样的化初创公司正在努力创建服务，以允许数据科学家也可以在该数据上训练他们的模型。
Got GPUs? Machine learning, and especially deep learning, require lots of expensive computations. Neural networks require the use of massively parallel computations, which GPUs are well suited for.
有GPU？ 机器学习，尤其是深度学习，需要大量昂贵的计算。 神经网络需要使用大规模并行计算，GPU非常适合。
Unfortunately, GPUs can be very expensive. But with tools like Google’s or Kaggle’s , anyone can run machine learning code in the browser using free (Tesla K80) GPUs.
不幸的是，GPU可能非常昂贵。 但是，借助Google 或Kaggle 类的工具，任何人都可以使用免费的(Tesla K80)GPU在浏览器中运行机器学习代码。
Algorithms are a commodity. Luckily for us, the machine learning field has built a culture of open source code and lots of result sharing. Whether its at the annual NIPS or ICLR conference, researchers tend to be very happy with sharing their results.
算法是一种商品。 对我们来说幸运的是，机器学习领域已经建立了一种开放源代码和大量结果共享的文化。 无论是在年度NIPS还是ICLR会议上，研究人员都对分享他们的结果感到非常满意。
If you want to keep up with the latest research, you can use the to read the latest papers in a beautifully indexed way. And there’s of course the machine learning . You can either use existing code, or use the free library to build your own models.
如果您想了解最新的研究，可以使用以精美的索引方式阅读最新的论文。 当然还有机器学习 。 您可以使用现有代码，也可以使用免费的库来构建自己的模型。
With great power comes great responsibility. You’ve got the code, you’ve got the data, you’ve got the computing power. Now you just need to be educated on how to use them!
拥有权利的同时也被赋予了重大的责任。 您有了代码，有了数据，就有了计算能力。 现在，您只需要学习如何使用它们！
Besides my channel of course, there are a ton of free educational resources out there to help you learn how to use the tools of machine learning. I made a three month machine learning curriculum that uses all of the above resources and several I’ve found on the Web to help a beginner get started. You’ll find that .
当然，除了我的频道外，还有大量免费的教育资源，可帮助您学习如何使用机器学习工具。 我编写了一个为期三个月的机器学习课程，该课程使用了以上所有资源以及我在网络上发现的一些资源，以帮助初学者入门。 您会在找到它。
You should be excited right now. This is an incredible time to be alive! There are so many changes happening so fast. Amidst all this complexity, machine learning can help us understand our world in ways we couldn’t otherwise. It can help us create and discover new things orders of magnitude more efficiently than ever before. You’ve got the power, use it wisely.
您现在应该很兴奋。 这是一个令人难以置信的活着的时刻！ 有这么多变化发生得如此之快。 在所有这些复杂性中，机器学习可以帮助我们以其他方式无法理解的世界。 它可以帮助我们比以往更有效地创建和发现数量级的新事物。 您有能力，明智地使用它。
翻译自:
机器学习与分布式机器学习
转载地址：http://jjewd.baihongyu.com/