It is hard to pursue a new learning or career path for many reasons. One for sure is the plurality of resources that you find online. Don’t get me wrong – this is fantastic! Just imagine the possibilities opening up, just because all that knowledge is available for everyone, for free…* In this blog post, you will find my learning resources and my comment on how they helped me develop. Other resources, recommendations are very much welcome!
AWS Training & Certification
Amazon offers various online courses around machine learning, as well as suggestions for learning paths. They were originally developed for Amazon’s internal use, until they decided to open them towards the outside world. Thanks for that! You can get AWS-certified after finishing their courses, but you don’t need to. I decided to roughly follow the Data Scientist learning path. When I started the first course on the path – The Elements of Data Science – the first sentence I heard was that ‘college-level mathematics’ were required for this course. Oh, boy. That is when I postponed starting this learning path, and go on a little detour to improve my maths skills. Later on, I rejoined the program and started my journey with the following courses.
That one was hard. I finished this course after the (much, much) more basic course on maths on edX.org (see below). It was very helpful that I already had the basic principles of maths from the previous course, and I was able to follow the argumentation and explanations nearly until the end. I had to give up on the derivation of some of the multi-dimentional matrices calculations, which I simply accepted (similar to my co-existence with maths throughout the years). The instructor talks really fast, but has a nice way of presenting to that I just had to repeat some of his explanations.
After the very hard introduction to basic mathematical principles used in ML, this one was a great relief. The content still is very challenging, and I took a lot of notes during this course. However, as I had done my basic work and learned Python as well as maths basics before, I was able to follow the explanations well. The course level in general was keeping the balance between challenge and fun. It also motivated to get started with machine learning – that is why I started my first ML project simultaneously with finishing this course!
The last time I came in touch with maths was in my last year at school. I have always been a good student, but I noticed early on that maths and physics were areas which didn’t come as easy to me as literature, languages or music. I had to get in touch with numbers and formulas again during my master’s degree, but maths and I always were happy when we just co-existed without having to fully understand each other (yes, I am sure maths didn’t get the gist of me, either!). A portal that I can hardly recommend to everyone – no matter whether you’re a beginner, intermediate or professional – is edx.org, where you can follow various online courses from renowned universities around the world. You can even get certificates for finishing the courses for a rather small amount of money (starting from around 40$ – it can get expensive though for certain programs). I enrolled in several courses during my journey.
The course is constructed for students who want to start a business degree and need a little refresher in maths before the semester starts. If you can vaguely remember your school days, this course is highly recommended. The examples are very easy and the course covers the most basic mathematical terms you will need. This course was important for me for another reason as well – as I went to school in Germany, my mathematical vocabulary in English was basically zero. I couldn’t even follow the most basic explanations because the terminology kept confusing me. In this course, I learned all the terms I needed. During later ML courses, I didn’t encounter any terminological problems, which made it much easier for me to follow the complex topics that were to come.
Although I did have basic Python knowledge, I still enrolled in this course as it seemed to be a great kick start in the direction of Data Science-related Python programming. And it was! The first two modules were a bit too easy (still, as they didn’t take long to finish, I went through with them), but the later modules were great for building up fundamental knowledge and revisiting what I had already known some months ago.
Corey Schafer’s YouTube channel is a great resource for hands-on Python tutorials (and not only Python – there’s a few other programming language tutorials, and he also has a very cute dog). I watched some of his videos for the more specific topics, for instance on how to work with APIs in Python. He has a concise way of presenting his knowledge, and his videos are easy to follow. They’re great as lunchtime entertainment. Although I always appreciate it if someone offers their knowledge ‘for free’, it is good to keep in mind that these people need to eat and have coffee, too – and to support them when using their services regularly.
Stack Overflow is the go-to address when I have a questions. It’s nearly 100% certain that there was someone before me with the same exact question, especially when starting with programming. A nice side effect of browsing these pages regularly is that you get familiar with how to post yourself – to using the search function first, to always provide sample output, your previous approaches and a thorough description of your problem, as well as the question you want to have answered. Trust me, it makes you a better person – not only when programming, but also in your business life. I was complimented several times during my time as a manager for my solution-oriented and thorough description of problems and issues – and I very simply just followed the basic posting rules that are used on Stack Overflow!
*I am aware that there are people who are not connected to the internet at all, and that it’s not completely free – given you have to use a technical device and have access to an internet connection.