Data science train is moving, at a constantly accelerating speed, and increasing its length by adding up new coaches. Businesses want to be on the data science train to keep up with the ever-evolving technology and improve their operations. Thus, there is a huge pool of jobs in the field of data science. More and more people want to get on the train and start working in this field because:

数据科学培训正在以不断加速的速度前进，并通过增加新的教练来增加其长度。 企业希望参加数据科学培训，以跟上不断发展的技术并改善其运营。 因此，在数据科学领域有大量的工作机会。 越来越多的人希望上火车并开始在这个领域中工作，因为：

1. The jobs are exiting and fun工作正在退出并且很有趣
2. The jobs are well-payed这些工作是高薪的
3. The demand will not decrease in the foreseeable future在可预见的将来需求不会减少

It is like a chain reaction. Businesses adopting data science create jobs that drive people to work in this field. These people need to be educated which motivates some other people to create learning resources. In this post, we will focus on the “learning resources” part of the story.

这就像一个连锁React。 采用数据科学的企业创造了推动人们从事该领域工作的工作。 这些人需要接受教育，以激励其他人创造学习资源。 在本文中，我们将重点介绍故事的“学习资源”部分。

The amount of resources to learn data science is overwhelming. There are two main reasons that cause this situation:

学习数据科学的资源数量巨大。 导致这种情况的主要原因有两个：

1. Data science is such a broad field that is kind of a mixture of math, statistics, and programming. Thus, there is so much to learn.数据科学是一个广泛的领域，混合了数学，统计和编程。 因此，有很多东西要学习。
2. People tend to prefer more flexible, faster, and cheaper learning paths over traditional education. Thus, there is a variety of MOOC courses, youtube videos, blogs, and bootcamps that teach data science.人们倾向于比传统教育更灵活，更快和更便宜的学习途径。 因此，有许多MOOC课程，YouTube视频，博客和训练营来教授数据科学。

There is so much to learn on so many different platforms. This can be an advantage or disadvantage depending on how we handle it.

在这么多不同的平台上有很多东西要学习。 根据我们的处理方式，这可能是优点还是缺点。

When I started learning data science, I had some questions that were demotivating me. I would like to list some of those questions here:

当我开始学习数据科学时，我遇到了一些困扰我的问题。 我想在这里列出一些问题：

• Should I learn Python or R? (I did not have any prior programming experience)我应该学习Python还是R？ (我之前没有任何编程经验)
• Do I need a masters degree or just a few certificates?我需要硕士学位还是只需要几个证书？
• How much statistics do I need to learn?我需要学习多少统计数据？
• I have a BS in engineering so I have enough math knowledge but “how much math do I need to learn” would be an important question for people with non-technical backgrounds.我拥有工程学学士学位，所以我拥有足够的数学知识，但是“对于非技术背景的人来说，“我需要学习多少数学”将是一个重要问题。
• TensorFlow or PyTorch?TensorFlow还是PyTorch？
• Should I learn natural language processing (NLP) techniques?我应该学习自然语言处理(NLP)技术吗？
• How about time series analysis?时间序列分析怎么样？
• What should I learn for data visualization? Matplotlib, Seaborn, Plotly or some other?对于数据可视化我应该学什么？ Matplotlib，Seaborn，Plotly还是其他？
• NumPy and Pandas enough for data analysis?NumPy和Pandas是否足以进行数据分析？

And there are some more questions. You might have similar questions and hesitate to start. I don’t have answers to those questions. But my suggestion is to stop wasting your time looking for answers.

还有更多问题。 您可能有类似的问题，开始犹豫。 我没有这些问题的答案。 但是我的建议是不要再浪费时间寻找答案。

Just start learning!

刚开始学习！

Once you start and take the first steps, you will discover some of the answers. You will also see that there is not a clear answer to some questions. However, this should not stop your learning process.

一旦开始并采取第一步，您将发现一些答案。 您还将看到对某些问题没有明确的答案。 但是，这不应阻止您的学习过程。

Another very important thing to keep in mind is that you cannot just learn everything. For instance, NLP is an entire field by itself and requires in-depth training and practice. If you want to specialize in NLP, you may focus more on NLP-specific tools and frameworks.

要记住的另一个非常重要的事情是， 您不能仅仅学习所有内容 。 例如，自然语言处理本身就是一个完整的领域，需要深入的培训和实践。 如果您想专门研究NLP，则可以将重点放在特定于NLP的工具和框架上。

There is always more than one option!

总是有不止一种选择！

When you start leaning towards a specific subfield of data science, some tools and frameworks become prominent but we usually have more than one option. For instance, R might be a better fit for statistical analysis than Python. However, Python also has powerful third-party statistical packages such as statsmodels. I’m not trying to cause any more contradictions. I just want to point out that there are many options to learn data science.

当您开始着眼于数据科学的特定子领域时，一些工具和框架会变得很突出，但是我们通常有多个选择。 例如，R可能比Python更适合统计分析。 但是，Python还具有强大的第三方统计软件包，例如statsmodels 。 我不是要引起更多的矛盾。 我只想指出，学习数据科学有很多选择。

I also want to mention different types of resources. It is actually good to have an overwhelming amount of resources. We have the freedom to choose from different options. There are videos on youtube about almost any topic related to data science. ArXiv contains a gigantic collection of scholarly articles on data science. Numerous platforms offer data science certificates such as Coursera, Udemy, and edX. And, of course, blogs are extremely efficient to learn specific topics. For instance, we can find an article on almost any topic on Medium.

我还想提到不同类型的资源。 拥有大量资源实际上是一件好事。 我们可以自由选择不同的选项。 youtube上有几乎与数据科学相关的所有主题的视频。 ArXiv包含大量关于数据科学的学术文章。 许多平台都提供数据科学证书，例如Coursera，Udemy和edX。 而且，当然，博客对于学习特定主题非常有效。 例如，我们可以找到有关Medium几乎所有主题的文章。

I started by completing a certificate, IBM Data Science Specialization. It was very helpful in the sense that the topics were organized and structured. It also provides a general overview of the field of data science. I suggest starting with a basic and comprehensive resource like that one. Then you will easily build your own learning path. You don’t have to collect lots of certificates on any topic.

我首先完成了IBM Data Science Specialization证书。 就主题的组织和结构而言，这非常有帮助。 它还提供了数据科学领域的一般概述。 我建议从这样的基础和综合资源入手。 然后，您将轻松建立自己的学习路径。 您无需就任何主题收集大量证书。

Last but not least, maybe the most important one, is doing projects. They are what get you ready for the job. Projects consolidate different skills into one. They also serve as a showcase to display your skills.

最后但并非最不重要的一点是，也许最重要的是做项目。 它们使您为工作做好准备。 项目将不同的技能整合为一个。 它们还充当展示您技能的展示柜。

Do projects after the basics are covered.

在介绍了基础知识之后再做项目。

Data science has lots of applications in different industries. The goal of businesses is to create value out of data. Thus, learning the algorithms or tools to analyze data is not enough to land a job. You should start doing projects related to the area you want to work in. Doing projects will not only help you obtain a more comprehensive knowledge but also bring more optimal tools and frameworks to your plate. Depending on the project, certain tools will outperform others and better fit to your style. Here is a list of my 5 reasons to do projects:

数据科学在不同行业中有许多应用。 企业的目标是从数据中创造价值。 因此，学习算法或工具来分析数据还不足以找到工作。 您应该开始进行与您要从事的领域相关的项目。进行项目不仅可以帮助您获得更全面的知识，而且可以为您的印版带来更多最佳的工具和框架。 根据项目的不同，某些工具的性能将优于其他工具，并且更适合您的风格。 这是我做项目的5个理由的清单：

To sum up, it doesn’t really matter how you learn. If you are passionate about learning data science, the path you follow does not make a difference. Whatever fits your learning style will do the job. The most important thing is to start your journey and, of course, do projects after you cover the basics.

综上所述，学习方式并不重要。 如果您热衷于学习数据科学，那么您所遵循的道路不会改变。 一切适合您的学习风格都可以胜任。 最重要的是开始您的旅程 ，当然， 在您介绍了基础知识之后再做项目。

Thank you for reading. Please let me know if you have any feedback.

感谢您的阅读。 如果您有任何反馈意见，请告诉我。