How do you fit data science learning into your routine?

I like to be able to vary my learning experience by mixing text book learning with articles, with podcasts and short videos. It allows the material to be explained from more than one perspective.

Setting a specific goal is the first the step for anything. Keep learning and sharing everyday is the second step. Earning certificates to authenticate yourself is the third step. Maintaining public profiles like LinkedIn, Kaggle or Medium etc is very important when you are done with your goals.

Think of it like planning a road trip: you wouldn't start without knowing your destination and stops along the way. Similarly, for data science, set small, digestible targets like "I'll learn linear regression this week," instead of vague goals like "I'll be an expert soon." This way, every time you hit a mini-goal, you get a boost of motivation, much like reaching a scenic viewpoint on your journey!

Setting realistic goals and maintaining consistency. Also, a certain degree of curiosity is important in the data science field.

Social media is flooded with resources and you can never reach the end of the abyss. So finding your calling is difficult. You have to run sample SPRINTS to FAIL FAST. Once you ascertain, you can create goals incrementally. Setting Goals was the easier part, but chasing them is harder. Two most detractors would be quality time and laziness. Believe me, I know. It is not how much time you spend but how you spend. My Gurus have told me to become "1% Better Everyday". I mean only incremental change can effect something substantial.

Choosing right resources is the important part of your journey and I think having a good mentor is plus the who can guide you to the right resources without going here and there. Medium, Kaggle are the best resources to learn everything about data science.

Data science is interdisciplinary. Once you get past the introductory stuff, lean away from learning data science from data scientists. Instead learn statistics from statisticians, software development and data management from software developers, data structures and algorithms from computer scientists, and science communication from science communication experts.

Youtube can be the greatest source for learning. However, it can also be something that can kill your time. Investing time in Selecting resources can be good, but deciding to learn via projects is better. Choose one delivery vehicle, say Python. My advice is to compound the power of Youtube resources and ChatGPT or BING or Bard to work on some project. Join communities in Kaggle. Additionally, help someone do their academic project.

I really liked the articles on Medium. Whether you are looking for high level information or in-depth, strategic or tactical they have it.

Allocate specific times in your day or week dedicated to learning. This could be reading articles, working through tutorials, or practicing coding. Consistency is key, so try to stick to your scheduled learning times as much as possible.

I think the best way is to choose right websites so instead of wasting time on social media, follow the right communities and spend time whenever you want to use social media. You cannot be regular all the time but if you follow good communities/groups then you can see every time something useful.

It’s easy to say “I’ll learn when I have time.” But often our existing work and tasks at hand consume all the time in our day, so we don’t get a chance to learn. I prioritize spending at least some time each day learning something new!

Schedule an hour daily in the mornings and enrol in some weekend data science courses to learn with a curriculum. Learning with a syllabus is important to make your journey structured

I don't like the idea of setting a schedule until it is a focus topic that I'm trying to gain more information on. I prefer to interact with the information on a daily basis as part of an ongoing conversation

If you want to learn it, you need to make the time. Once you have laid out your goals and defined your resources, you need a plan of attack. Don't be afraid to use tools like ChatGPT to help you build your learning plan. I take one or two lunches a week to watch a video on YouTube or LinkedIn Learning. This has been helpful in getting in some quick learning for a hot topic (e.g. GenAI, vector and knowledge graphs, etc). These could be an ad-hoc learning on top of your focused structured plan.

Offering to teach or share your project (and its outcomes) with others is one approach to try out at least once! It makes us both learn the topic well and then try to break it apart and explain it to an audience in a way that they can understand.

There are volunteer opportunities to try out your learnings. For example, Omdena projects, Data for Good org, and World AI Hackathons provides several legitimate projects for aspirants to work on. Each of these projects may take time and hence its important to allocate specific if you were to take these projects.

Enroll in a weekend program where you are given projects to apply your learning. Kaggle is another platform to apply learning and there is a huge opportunity.a

Not only should you set goals but you should also set some milestones. These will help you check your progress which is important for your self-assessment. Take an agile approach and iterate on what is working well and what isn't and then adjust. Python projects too hard? Put a pause on the that course and take an intro to python course then circle back for example. Keep organized notes on topics and tag ones that you need to revisit for clarity. This has helped me out a lot. I use Obsidian so I can tag and link notes together which has helped me immensely.

Find a topic that you already care about. For example, if you're passionate about sports, learn from a sports dataset instead of the (overused) Titanic and Iris datasets. You will find that your subject matter expertise will help you better questions about the data, make hypothesis, spot errors and keep you motivated. More importantly, if this learning project ends up in your porfolio, it may help you get a job in an area that you are already passionate about.

Learning is very similar to exercising. Having “fun” while learning is the best way to enhance the continuity of learning. It does not matter if you learn a big ML algo or a small function in any analytical tool; the key is to learn and do it on a continuous basis. Learning can be from books, online docs, codes written by peers or by just interacting with your colleagues.. there is no right way or wrong way but choose a way that you enjoy and are likely to stick to.

It’s interesting that for me this last step is usually my first. I have learned about myself that I am more likely to complete learning a new topic when curiosity sparks. When it is a must know topic one must learn and it may not necessarily be the most fun, understanding the importance of why the topic is important to learn is also a good motivator for me. Finding time to learn might be a challenge too, For me, the best time for learning is in the morning. You may have an ideal time to learn, perhaps it’s in the morning commute, after work, or depending on how flexible your time during the day may be perhaps time blocking may work for you. Socializing what you have learned is also a great way to normalize learning in any environment.

Agree that staying curious on the topic and asking the questions can keep your interest alive and may lead you to explore other related topics.