机器学习与分布式机器学习_我如何学习机器学习

机器学习与分布式机器学习

My experience on How I got started and gradually learned the basics.

我在开始和逐步学习基础方面的经验。

“The best way to learn Machine Learning is by DOING IT”.If you understand what I mean by this statement, then there is probably no need to go through the rest of the article.

“学习机器学习的最佳方法是通过DOING”。如果您理解我的意思，那么可能不需要遍历本文的其余部分。

When I was starting, I didn’t know where to begin with. Too much information, too many courses, and just too many different varying opinions on what to do and what NOT to do.

当我开始的时候，我不知道从哪里开始。 太多的信息，太多的课程以及关于做什么和不做什么的太多不同意见。

I will be honest, when I googled How to start learning machine learning(or something like that)for the first time back in 2019, I was bombarded with a plethora of varying opinions and different ways to get started.

老实说，当我在2019年首次谷歌搜索如何开始学习机器学习(或类似的东西)时，我受到了各种各样的观点和不同的上手方式的轰炸。

Some said to learn R, Some said to start with Python, Others recommended getting a Masters degree, while a few downright told me to take a different career path. Yea, I know, varying opinions.

有些人说要学习R，有些人说是从Python开始的，另一些人则建议获得硕士学位，而另一些人则要求我采取不同的职业道路。 是的，我知道，意见不一。

All of this led to an Information overload and left me more confused than ever on where to begin with.

所有这些导致信息超载，使我从头开始比以往更加困惑。

BUT I knew I had to start somewhere and I did.

但是我知道我必须从某个地方开始，但是我做到了。

Made a ton of mistakes and still make them to this day but mistakes are good if you learn from them and make yourself better.

犯了很多错误，直到今天仍会犯错，但是如果您从错误中学习并使自己变得更好，那么错误就很好。

It’s an iterative process which makes you better along the way.

这是一个反复的过程，可以使您不断进步。

So without further delay, here are is my step by step guide on how to get started with Machine Learning:

因此，不用再拖延了，这是我的机器学习入门指南：

从统计开始 (Start with Statistics)

Photo by Stephen Dawson on Unsplash

Stephen Dawson在 Unsplash上 拍摄的照片

Learn Statistics, Clear your concepts around statistics especially pertaining to machine learning algorithms.

学习统计信息，清除有关统计信息的概念，尤其是有关机器学习算法的统计信息。

When I was starting out, I thought my job would be confined to “programming” only and that there was no need to dive deep into the mathematics as I had Libraries to take care of that. Clean the data, fit a Linear Regression Model, and job well done.

当我刚开始的时候，我认为我的工作将只限于“编程”，并且因为我有图书馆来照顾这个问题，所以没有必要深入研究数学。 清理数据，拟合线性回归模型，并做好工作。

Well, I was WRONG.

好吧，我错了。

A clear understanding of Statistics is a MUST and I would like to give an example of why it is.

必须对统计数据有一个清晰的了解，我想举一个为什么这样的例子。

When I was practicing with Linear Regression, I read somewhere that Outliers can pose a problem to it but I did not know how to detect and deal with them.

在使用线性回归进行练习时，我读到某个地方的异常值可能会对它造成问题，但我不知道如何检测和处理它们。

That’s where I found out that z-score and Inter-Quartile Range, two very important statistic concepts, are used exactly for that, to detect Outliers.

那就是我发现z-score和四分位间距(Inter-Quartile Range)这两个非常重要的统计概念的目的，正是这些用于检测异常值。

And in another example, I learned How P-values and null hypothesis are extremely important to detect insignificant variables present in our dataset.

在另一个示例中，我了解了P值和零假设对于检测数据集中存在的无关紧要的变量非常重要。

And this is just scratching the surface.

而这只是表面。

So in short, get good with statistics and the book I used to learn the basic concepts is : Introduction to Statistical Learning.

简而言之，请精通统计知识，而我用来学习基本概念的书是： 《统计学习入门》 。

熟悉Python (Get comfortable with Python)

Photo by Hitesh Choudhary on Unsplash

Hitesh Choudhary 摄于 Unsplash

Learn Python, Simple as that. There are a ton of resources out there that provide extensive content when it comes to Python. A lot of people recommend R as well but the thing is, You can’t learn everything and there is no need to learn different tools which perform the same function. What R can do, Python can do just as well if not better. So instead of dedicating significant time and effort into learning two languages, Focus on mastering one and down the road when you do get the time, Learn R too if you feel like it.

学习Python，就这么简单。 关于Python，有大量资源可以提供广泛的内容。 很多人也推荐R，但事实是，您不能学所有，也无需学习执行相同功能的不同工具。 R可以做什么，Python甚至可以做得更好。 因此，与其花大量的时间和精力来学习两种语言，不如在有空的时候专注于掌握一种语言，如果愿意，也可以学习R。

A great place to start learning Python is this tutorial.

本教程是开始学习Python的好地方。

开始练习数据集 (Start practicing on Datasets)

Photo by Markus Spiske on Unsplash

Markus Spiske在 Unsplash上 拍摄的照片

Once you have extensively worked on the first two tips, then move onto Kaggle and start implementing what you have learned. I cannot emphasize this point enough.

在对前两个技巧进行了广泛的研究之后，请继续学习Kaggle并开始实施所学知识。 我不能足够强调这一点。

There is no point in learning a concept if you ‘re not going to implement it.

如果您不打算实施一个概念，那么学习它是没有意义的。

Use your newly learned coding skills to clean the data, create wonderful visualizations, and fit Machine Learning models. Understand the data and the behavior of the model. Like for example You know, Outliers are bad. Well, Why are they bad? What effects do Outliers have on our model? Be an investigator and be curious, very curious.

使用您新近学习的编码技能来清理数据，创建出色的可视化效果并拟合机器学习模型。 了解数据和模型的行为。 例如，您知道，异常值是不好的。 好吧，为什么它们不好？ 离群值对我们的模型有什么影响？ 做一个调查员，好奇，非常好奇。

Recommnded dataset to get started with:

推荐的数据集开始：

Classification :Titanic Dataset, Iris Dataset

分类：泰坦尼克号数据集，虹膜数据集

Regression: Boston House pricing , Auto Dataset

回归：波士顿之家定价，自动数据集

获得真实的生活经验(Get real life experience)

Photo by Ian Schneider on Unsplash

Ian Schneider在 Unsplash上 拍摄的照片

So you now know the basics of statistics, you can code and you have a few Kaggle projects on your portfolio as well. congratulations, you’ve made it. Good job, Well done. BUT I hate to break it to you that You are just getting started.

因此，您现在知道了统计的基础知识，可以进行编码，并且投资组合中也有一些Kaggle项目。 恭喜，您成功了。 干得好，做得好。 但是，我不想让您知道您才刚刚起步。

I already went over this point extensively in one of my previous Articles where I mentioned how an internship was an eye-opener for me.

在之前的一篇文章中，我已经广泛地谈到了这一点，其中提到了实习对我来说是大开眼界的。

Where I for the first time got the chance to provide Real Value in terms of business to Real People.

我第一次有机会向Real People提供业务方面的Real Value。

Real Life job experience will humble you, pressurize you, break you BUT will teach you A LOT if you are the type who is hungry for learning.

现实生活中的工作经验会让您感到沮丧，压力，折磨，但如果您是渴望学习的人，那么它会教给您很多。

Conclusion

结论

I am no different than anyone reading this article, a student of the field of machine learning who is hungry to get better every single day. As of now, I am working as a full time internee which is quite humbling. I am still learning and my journey has just started.

我和阅读这篇文章的人没有什么不同，我是机器学习领域的一名学生，渴望每天都变得更好。 截至目前，我正在做一个全职的实习生，这真是令人感到谦卑。 我仍在学习，我的旅程才刚刚开始。

NOTE: If you like my writing and the content I post, feel free to share it with your friends and family for that helps me a lot. Thank you :)

注意：如果您喜欢我的文章和我发布的内容，请随时与您的朋友和家人分享，这对我有很大帮助。 谢谢 ：)

翻译自: https://medium.com/swlh/how-did-i-learn-machine-learning-e72eb151afd3

机器学习与分布式机器学习