The Biggest Mistake I Did When Learning Data Science Which took me a long time to realize

I would like to start with stating a ground truth just in case you have not realized by now: Data science is an extremely broad field.

Data science can be applied to any business or industry where we can collect data. Besides, the advancements in data-related technology has made it easier than ever to collect, process, store, and transfer data. Thus, it is safe to say that data science applications will cover a broader scope in the future.

Although data science is ubiquitous, its applications differ greatly in different domains. It would be an uphill battle to learn about all data science applications. I now consider it as a battle that is impossible to win.

It is in your hands to turn this battle into an achievable target. You should set your goal towards being a specialist rather than a generalist. This is how you make great impact as a data scientist.

Being a jack of all trades but master of none will get you only to a certain point. What you can achieve at that point will not impress potential employers . You should try to become a master of at least one area to stand out.

The learning path mostly starts from the same point for all domains. The core principles are the same. You need to have a comprehensive understanding of statistical concepts. A certain level of programming skills is needed to turn your ideas into action.

Once you are through with the basics, pick an area to specialize. You do not have to work in the first area you pick. You can always change it. In fact, while obtaining in-depth knowledge in a particular field, you also improve your data science skills and knowledge in general.

Let’s say you want to work as a data scientist in finance. Then, time series analysis should be your area of expertise. You need to be able to clean, process, and analyze time series data as well as extract insights from it.

How about tools?

The jack of all trades but master of none is also a serious issue in terms of software tools and packages. Thanks to the data science community, we have a rich selection of tools that ease and expedite our jobs.

The advantage of having so many tools may turn into trouble if used unwisely. There are almost always more than one option to perform a task.

Consider a very simple case of cleaning and analyzing tabular data. The first two options that come to my mind are Pandas for Python and data table for R. SQL is also another strong candidate especially if the data is stored in a relational database.

Similarly, there are many candidates to help you with data visualization tasks. Matplotlib, Seaborn, and Altair are just the three options in Python.

In most cases, one is enough to get the job done. You won’t be at a disadvantage because of using Seaborn instead of Matplotlib, and vice versa.

The hardest decision here is to pick a subfield of data science to specialize. Unfortunately, there is not a strict set of rules to help you make this decision. It depends on many parameters such as your background, interests, and job opportunities.

Whatever area you pick, it will be better than learning about all. You are highly likely to fail with the latter. What I mean by fail is not that you cannot learn anything. However, you will fail to impress recruiters with your general knowledge or skills. Learn More

Thank you for reading. Please let me know if you have any feedback.