机器学习1000问_您一直想问的机器学习问题

经验和建议 (Experience and Advice)

介绍 (Introduction)

At times pursuing a career or some form of qualification within the machine learning field can get very difficult.

有时，在机器学习领域追求职业或某种形式的资格认证会变得非常困难。

Most of these difficulties stem from having to make decisions that can alter the course of your professional or academic career, so you want to make sure you are making the right choice.

这些困难大多数来自必须做出可以改变您的专业或学术职业路线的决策，因此您需要确保做出正确的选择。

Should I go to University A or University B?

我应该去大学A还是大学B？

Should I become a Computer Vision Engineer, or go down a pure Data Science route?

我应该成为计算机视觉工程师，还是走纯粹的数据科学路线？

Questions such as the ones above are ones you might be asking, and they are ones that machine learning practitioners of all levels have asked me via LinkedIn.

诸如此类的问题是您可能会问的问题，并且是所有级别的机器学习从业者都通过LinkedIn询问的问题。

LinkedIn is a communication channel that transcends physical distance, and it has grown in popularity over the years. It has become a social network space for professionals.

LinkedIn是一种超越物理距离的沟通渠道，并且多年来已经越来越流行。 它已成为专业人员的社交网络空间。

在本文中，我将包括机器学习从业人员和学生提出的一些常见问题，并且还将包括我提供的相应答案。 (In this article, I will include some common questions that have been asked by machine learning practitioners and students, and I’ll also include the corresponding answers I’ve provided.)

I have removed any names and personal information from questions to protect the identity of individuals.

我已从问题中删除了任何姓名和个人信息，以保护个人身份。

Disclaimer: All answers provided are based on my experience, and I always advise individuals to make decisions that best suit their current situation and to go further and conduct research.

免责声明：提供的所有答案均基于我的经验，我始终建议个人做出最适合其当前情况的决定，并进一步进行研究。

问题1(以前的研究经验) (Question 1 (Prior research experience))

Hi! I came across some of your Medium articles about pursuing an MSc, I was wondering did you have much research experience in order to get admitted into grad studies? Thank you, your articles were a great read!

嗨！ 我在您的中型文章中找到了一些有关攻读MSc的文章，我想知道您是否有很多研究经验才能入读研究生课程？ 谢谢，您的文章读得很好！

回答 (Answer)

Hi, thanks for reaching out to me and reading my articles.

嗨，谢谢您与我联系并阅读我的文章。

In regards to how much research experience I had in order to get admitted to grad studies: I had no academic research experience.

为了获得研究生学位，我需要多少研究经验：我没有学术研究经验。

Before going for my MSc in Computer Vision/Machine Learning, I only had a BSc in Software Engineering and two years working as a Web developer.

在攻读计算机视觉/机器学习理学硕士学位之前，我只有软件工程理学学士学位，并曾担任Web开发人员两年。

Although I had conducted some independent exploration of the machine learning field and even bought some textbooks, after a few months of self-study, I decided to go for an MSc in order to commit myself to study machine learning and other related topics thoroughly.

尽管我对机器学习领域进行了一些独立的探索，甚至买了一些教科书，但经过几个月的自学，我还是决定去读MSc，以致力于研究机器学习和其他相关主题。

I hope I’ve answered your question, feel free to ask any further questions.

希望我已经回答了您的问题，随时提出其他问题。

问题2(先前的研究经验) (Question 2 (Prior research experience))

Recently I’ve been thinking about grad school for ML and the universities I’ve looked into all ask for research experience. Did you find it difficult to get into grad school without having done formal research or did you find that your personal projects/ self study were sufficient?

最近，我一直在考虑学习ML的研究生院，以及我研究过的所有大学都在寻求研究经验。 您是否发现不进行正规研究就很难进入研究生院，还是发现自己的个人项目/自学能力足够？

回答 (Answer)

Getting into grad school wasn’t difficult for me as I met the prerequisite requirements the University set for the course.

对我来说，进入研究生院并不困难，因为我达到了大学为该课程设定的先决条件。

Most University tend to set out background knowledge that students applying for the course should have.

大多数大学倾向于列出申请该课程的学生应具备的背景知识。

For my particular course having a research background wasn’t necessary at all, but having a computer science background and experience with programming was crucial to obtaining a place on the course.

对于我的特定课程，根本没有研究背景，但是拥有计算机科学背景和编程经验对于在课程中占一席位至关重要。

Some universities might require a research background if you are applying for PhDs.

如果您申请博士学位，有些大学可能需要研究背景。

My grad school degree was at the Masters level. Undertaking a post-doctoral degree does entail a lot of research effort from students, so it’s very understandable as to why Universities will only admit students that have prior research experience.

我的研究生学位是硕士学位。 攻读博士后确实需要学生进行大量的研究工作，因此，为什么大学只招收有过研究经验的学生，这是可以理解的。

I would advise that you should examine the prerequisite requirements set out by the University you are planning to attend and see if you meet them, even talk to faculty members at Universities you would like to attend for further advice.

我建议您应该检查您计划参加的大学提出的先决条件，看看是否满足要求，甚至可以与您想参加的大学的教职员工交谈以获取进一步的建议。

问题3(机器学习专业) (Question 3 (Machine Learning Speciality))

My question is what areas of machine learning do I need to focus on for my master’s degree

我的问题是，硕士学位需要关注哪些领域的机器学习

回答 (Answer)

A simple answer would be to work on the area that you are most interested in, but also focus on any areas within machine learning that you might be lacking in.

一个简单的答案将是在您最感兴趣的领域上工作，但还要专注于您可能缺乏的机器学习领域。

From my experience, before embarking on my MSc, I knew my maths knowledge was below average, and I needed to learn some key maths topics before the course started. So I took some crash courses on Linear algebra, Statistics, Calculus and Mechanics etc.

根据我的经验，在修读MSc之前，我知道我的数学知识还没有达到平均水平，因此我需要在课程开始之前学习一些关键的数学主题。 所以我修了一些关于线性代数，统计学，微积分和力学等速成课程。

In terms of following my interest, I focused on the Computer Vision aspect of Machine Learning because I prefer working with videos and image data as opposed to numerical or text. Also, Deep Learning at the time was very prominent, especially with the advent of commercial self-driving cars.

在关注我的兴趣方面，我专注于机器学习的计算机视觉方面，因为我更喜欢使用视频和图像数据，而不是数字或文本。 此外，深度学习在当时非常杰出，尤其是在商用自动驾驶汽车问世时。

Some individuals focus on areas such as Natural language processing(NLP), Data Science, Speech and audio Recognition, Medical Imaging etc.

一些人专注于自然语言处理(NLP)，数据科学，语音和音频识别，医学影像等领域。

My advice would be to learn a bit of the common areas within Machine learning. After getting a grasp of ML as a whole, you could then specialize in the areas where you are more interested in, or where there are more job opportunities.

我的建议是学习机器学习中的一些常见领域。 掌握了ML的整体知识之后，您便可以专注于您更感兴趣的领域，或有更多工作机会的领域。

Hope this was helpful, feel free to ask any more questions

希望这对您有所帮助，请随时提出其他问题

问题4 (Question 4)

I‘m working as iOS developer, Can I find an application that makes use of machine learning in iOS development or something like that?

我正在作为iOS开发人员，我可以找到一个在iOS开发中利用机器学习或类似功能的应用程序吗？

回答 (Answer)

In my current role as a computer vision engineer, I develop machine learning models for iOS-based application. So I’ve implemented solutions for pose estimation, semantic segmentation, gesture recognition and face detection for iOS application.

在我目前担任计算机视觉工程师的职位上，我为基于iOS的应用程序开发了机器学习模型。 因此，我为iOS应用程序实现了用于姿势估计，语义分割，手势识别和面部检测的解决方案。

Having machine learning models running on smartphones and other edge devices is very relevant.

在智能手机和其他边缘设备上运行机器学习模型非常重要。

Instagram, TikTok and Snapchat all use machine learning models optimized for mobile devices, so it’s definitely an area worth exploring.

Instagram，TikTok和Snapchat都使用针对移动设备优化的机器学习模型，因此，这绝对是一个值得探索的领域。

结论 (Conclusion)

I hope that you have found some value from the content within this article.

我希望您从本文的内容中发现了一些价值。

It’s very humbling that there are people that view my experience and expertise as a learning point. If you have any questions that you would like to ask me, or perhaps you would prefer if I elaborated on answers to some question in more detail, then you can reach me through LinkedIn as usual.

有人将我的经验和专业知识视为学习要点，这真是令人感到羞耻。 如果您有任何疑问想问我，或者您更希望我详细说明某个问题的答案，那么您可以像往常一样通过LinkedIn与我联系。

I am not reluctant to answering machine learning related questions or queries as I know how hard and challenging the field can be, so please don’t be shy to ask any pressing questions. I’ll try my best to provide suitable answers.

我不愿意回答与机器学习相关的问题，因为我知道该领域可能会多么艰巨和具有挑战性，所以请不要害羞地提出任何紧迫的问题。 我会尽力提供适当的答案。

我希望您觉得这篇文章有用。 (I hope you found the article useful.)

To connect with me or find more content similar to this article, do the following:

要与我联系或查找更多类似于本文的内容，请执行以下操作：

1. Subscribe to my Email List for weekly newsletters

   订阅我的电子邮件列表以获取每周新闻

2. Follow me on Medium

   跟我来中

3. Connect and reach me on LinkedIn

   在LinkedIn上联系并联系我