How to achieve your learning goals for data science

As a self-taught data scientist or wanna-be data scientist, deciding on what to learn per time is important. However, it can become a stressful task in itself because there a lot of materials and new concepts to cover. For some people, the challenge is that they lack the motivation to keep learning. They quit after a week or two and then start all over again and have been going in this cycle ever since.

So I have written this article for you, if:

• you feel like you aren’t learning fast enough or learning as much as you would like to
• you are easily distracted by new courses or topics coming out of the field of machine learning
• you are a complete newbie and you don’t know where to start
• you need a structured learning approach with achievable goals in 2019.

Truth is, we would encounter new interesting resources, courses or frameworks along this journey. And the challenge we’d face is NOT to get carried away by them – the shiny new courses on XYZ framework or the latest book by an established data scientist on a new area of machine learning. While I’m not against staying up to date with new resources, they can be a form of distraction that gets us nowhere. So prioritizing our time as we learn, and making progress is of essence for a self-taught data scientist. Else, we’d find ourselves jumping all over the place trying to learn about the latest topic on Reddit and Twitter every single day.

Begin with the end in mind

In his book, The Creative Curve, Allen Gannett says there are four laws that we can all follow to master any skill:

• Law of Consumption – reading books on the topic, watching tutorials and listening to podcasts etc on the topic. He recommends spending 20% of our waking hours consuming a lot of materials on the subject. You don’t just consume, you seek to understand the material. You dig deep into the concepts. That’s consumption!
• Law of Imitation – copying the steps taken by the author, redoing the tutorials as we have watched it and rewriting the code from scratch to reinforce our understanding. Imitation gives you the hands-on experience and the technical know-how on the materials you have consumed. You imitate to find patterns and reinforce the materials you have consumed. This was how Ben Franklin learned to write.
• Law of Creative Community – asking questions from those who have done it before. Writing the author of those materials, joining an active community on Reddit, Discord, Slack or Twitter. Seeking a mentor. Finding an accountability partner.
• Law of Iteration – applying what we have learned to solve a problem. This could be a small project, exercises from textbooks, Kaggle competitions or fixing bugs. Iteration brings all you have done together and makes you carve your own niche and make your own project.

So as we start the new year I want us to stay grounded in our journey. Ask yourself: what one thing do I want to achieve in January? Then, what are the tools or resources that will help me achieve this goal?
A month is a short time in our learning journey. However, we can use it as a personal metric to quantify our achievements and measure progress. For instance, lets say you are still a beginner who is already familiar with the concepts and tools used in Machine Learning and you want to answer the questions I asked earlier: what one thing do I want to achieve this month? I want solve a Kaggle problem this month.

Then, what tools or resources will help me achieve this goal? Here is a sample plan with resources.

This is a good plan with concrete steps. It’s very important to be super specific so please go ahead to include the time of the day you want to do tasks or change how much time you would commit to each one of them. I’d advice that you keep the tasks not more than six so you don’t get bugged down with a lot of them.

Now that you have set the tone for the month of January, you begin each day with the end in mind. Your mind focuses on the learning tasks and how you can do them. Your thoughts are directed at replicating the codes and digesting the concepts you have studied – even while at work. You get to your desk to study in line with the plan. You don’t worry about implementing the latest paper on Convolutional Neural Network. You don’t spend 3 hours trying to research the difference between Pytorch and Tensorflow because of the conversation you saw on Twitter. You don’t have FOMO (Fear Of Missing Out) on Generative Adversarial Networks because its generating a lot of buzz on Reddit. You’re keen on following through with your plan. I know setting goals alone may not guarantee that you would do all the tasks above – especially if you are very busy. But I can guarantee that it would help you feel much more accomplished as you continue to carry out these tasks. It will also help you refocus your mind whenever you are distracted by social media and what’s hot or trendy. Ultimately as a data scientist, your job will be to prioritize what project(s) are important, so this structured approach is a good start in learning to prioritize your workflow – a useful skill every data scientist needs.

Take Action

So I challenge you as you approach the new year to look at each month as a canvas where you can write out your learning goals and actually do them. Think critically about the area or topic(s) you want to skill up on or a project you want like to work on. Write out 5 or 6 concrete and detailed steps that would help you achieve it and put it on your sticky note on your computer or where you’d see it every single day. Do this every month, refine your plans when its needed but stick to it.

Having concrete goals help you prioritize your decisions and time.

If this is your first month trying to become a self-taught data scientist, I suggest you check to see the study plan I created that got me started on this journey, see here.

Keep Learning!