What is the most important trait/skill a Data Scientist must have? This question does rounds all over the information technology world these days. And answers may be varied. Some may say math & stats others programming. Matured ones may say business knowledge, curiosity, problem solving, etc. All of them are right in their own place. But underlying them is a key trait; the ability to think from First Principles.
First principle thinking in Data Science revolves around two questions:
- What are we trying to achieve?
- Why do we want to achieve it?
A classic rookie mistake is that they skip the above and move to the ‘How’. For instance, which algorithm to use? how to train a model? how to evaluate? But they forget to ask, “What is the problem to be solved?” , “Why do we need ML?”
Being a Data Scientist is like being a warrior. You need to understand what weapon to use in a scenario. As we have said in an article of ours:
A Data Scientist is the one who not only knows how to use machine learning, but the one who knows when to avoid it.
Benefits of First Principle Thinking for a Data Scientist
Nonetheless, how does first principles thinking help Data Scientist? For starters, it helps you with clarity of scope. It helps you understand what not to do, when a google search gives you hundreds of solutions.
Secondly, it helps you come up with creative solutions. In a fast-paced world with exponential technologies and everyone trying to be ahead in the race, it’s rare to find solutions that are simple yet elegant.
Third, it helps you communicate effectively. Being a Data Scientist, you regularly communicate with stakeholders. Being clear with first principles helps you communicate effectively and get the business stakeholders onboard.
Conclusion
These are opinions based on personal experiences. This is not intended to discredit other thoughts, which are equally valid. Moreover, we do not claim any guarantees regarding the completeness or accuracy of these ideas.
Also Read : Elements of a Data Science practice
Inspired from a twitter thread of mine.
