Why I Write a Data Science BlogBlog writing has accelerated my learning in data science, here’s whyRebecca VickeryBlockedUnblockFollowFollowingMar 23Photo by Kaitlyn Baker on UnsplashA little over 7 months ago I wrote my first data science blog post on Medium.
I have since written numerous posts sharing resources, tutorials and my opinions on learning data science.
When I first started my intention was simply to share my learning journey with others in the hope that I could provide some useful information for other people embarking on a similar path.
However, over the last few months I have experienced some really vital benefits to writing about data science, and found that it has accelerated my learning.
In the following post I wanted to cover some of the key benefits I have experienced:Feynman techniqueIn the early days of learning data science, and even now I often find myself copying and pasting bits of code from various sources, and using it to solve a problem, without fully understanding how it works.
This is fine and there is definitely a place for this in allowing me to get something done quickly.
But I have found that writing a blog about something new that I have learnt allows me to fully understand it.
For example, if I am writing a tutorial then I have to understand exactly how and why something works in order to explain that in the blog.
I also end up researching a lot of the theory behind it which expands my knowledge of data science.
This I think is similar to the Feynman technique.
This is a mental model for learning developed by physicist Richard Feynman.
The basis of this method of learning is that in order to fully master a subject you need to be able to explain it back to somebody in simple terms.
When I use or learn a new data science technique I will write a post about it.
This allows me to consolidate my understanding by explaining it back.
Writing a blog post helps me to fully grasp a concept.
Public critique of workI have largely studied data science through self led learning so I have not necessarily had teachers to ask questions to or to check that my understanding is in fact correct.
The internet, however, is a great place to get feedback.
When I write something into a blog post, particularly if it contains code, then if there are mistakes or if something isn’t quite right, the chances are that someone will leave a comment.
It is much better to make the mistakes when working on a toy problem, and correct them, than to make the mistake when using the technique in a workplace scenario.
“Writing and learning and thinking are the same process.
”, William ZinsserConnecting with othersWriting my blog has enabled me to connect with others who are following a similar learning path, or who are already working in the field of data science.
I have found in particular writing on Medium, as it is to some extent a social media platform as well as a blogging platform, facilitates this well.
It has been really useful to find that there are others out there following a similar learning path and to connect with them.
CommunicationOne of the most important skills needed to work as a Data Scientist is communication.
You need to be able to explain complex technical concepts to many different audiences and to be able to tell stories with data.
At work I might have to explain how a machine learning model works at many different levels, from fellow data scientists to a business sponsor, or executive, who has no technical knowledge of the process.
Writing a blog has been great practice — it has enabled me to organise my thoughts, write a walk through of a workflow and just generally get more practice at writing which makes it faster for me to create written reports.
Store of workI have found that my blog posts have become a great store of code snippets and information.
I often need to use a technique again at a later date by which time I have forgotten either the code I need to write or a particular nuance around how it works.
I find that I regularly revisit my blog posts to take pieces of code to use in my work.
There are of course other places to store code to access later, but a blog has the added advantage that the code is highly annotated.
I have really seen an acceleration in my learning, a deeper understanding of data science techniques and a vast improvement in my communication skills since starting my blog.
I really didn’t realise just how useful writing could be as a learning technique, and of course it comes with the added benefit of being able to provide a store of information for others to also use.
.. More details