Learning how to learn is one of the most useful skills you can cultivate.
When I first started teaching myself programming and data science in 2018, I enrolled into countless online courses. Every time I completed a course and got a certificate, I’d get a momentary feeling of accomplishment.
As though I’d learnt something new.
That feeling, however, was always fleeting, because every time I tried to put what I learnt into practice, I failed.
Theoretically, I understood how classes, methods, and object-oriented programming worked. I knew the difference between random forests and decision trees, and grid search versus Bayesian optimization.
Yet, I wasn’t able to work with real datasets.
Every time I tried to build a new project, I would come across a roadblock and encounter yet another topic I didn’t know about.
Then I’d go down a rabbit-hole of taking new online courses to fill the gaps in my knowledge.
This endless cycle is known as the tutorial-trap, and I was stuck in it for two years.
Until I learnt how to learn.
After speaking to programmers and data scientists who already had jobs in the field, I curated a data science roadmap that I followed to teach myself the subject.
In just a few months, this roadmap helped me land a job in the field.
I used similar roadmaps to climb the corporate ladder, improve my communication and data storytelling skills, learn new languages like Pyspark and SQL, write engaging articles on Medium, and create my first data science online course.
All the learning roadmaps that I’ve created previously, however, were applied before ChatGPT was released last year.
When I transitioned into a new role in my company that required me to build time-series forecasting models, work with data visualization frameworks that I haven’t…