The Reason You’re Frustrated when Trying to Become a Data Scientist

• Read original article here

Photo by Sebastian Herrmann on Unsplash
The Reason You’re Frustrated when Trying to Become a Data Scientist
The hidden skill that separates the best from the rest
Jun 16 · 6 min read
Over-demanding Expectations…
How many times have we seen the post “5 things you need to become a Data Scientist”, “How to become a Data Scientist in 2020”, or the images of the Venn diagrams?
Don’t feel bad if when you read the requirements you curled up into a ball, sucked your thumb and procrastinated even harder on your goals because it’s unlikely you’re alone in this situation. If you are frustrated, its arguably not entirely your fault as to why you are feeling this way.
Data Science is a large field with many cross sections to other disciplines, but I think we have complicated the criteria for becoming a Data Scientist with many complex prerequisites, which are required further down the line, but are not what will keep you going in the long run.
Anyone can become a Data Scientist. It takes is the will to do it and the desire to carry out whatever it takes. Two traits of which every human can realize.
Am I saying that you do not need to know some key topics such as Linear Algebra, Statistics, Calculus, a Programming language – Python and R seem to be the most popular – and heck load of other stuff? Of-course not! To understand the inner workings of Logistic Regression or to dissect a typical research paper, you are probably going to want to know some linear algebra, or statistics or calculus — depending on what paper you are reading — if you want to make it out alive.
What I am saying is that the fundamental skill that is required to become Data Scientist (and to remain one for the long haul) is not Data Visualization, understanding Machine Learning algorithms, or all the others that we I’ve already listed, and those that we have been told that I have not added. Instead, the fundamental skill is the ability to learn, quickly!
“The illiterate of the 21st century will not be those that cannot read and write, but those who cannot learn, unlearn and relearn” — Alvin Toffler
I have written a story about the 3 stages of Learning Data Science that discusses the process of learning which, when understood, can boost how fast you learn.
towardsdatascience.com
Defining the terms
Learning is the process of acquiring new understanding, knowledge, behaviors, skills, values attitudes or preferences. We regard someone to be a good learner when they are consistently updating the aforementioned things effectively in a manner that betters their well-being.
Data Science is an inter-disciplinary field that uses scientific methods, processes, algorithms and systems to extract knowledge and insights from many structural and unstructured data (source: Wikipedia ). An effective Data Scientist is consistently capable enforcing techniques that constantly allows them to extract knowledge and insights that could be used to solve real world problems.
The current state of affairs
The current issue that I believe we have when someone wants to begin their journey into Data Science is similar to the problem of the current academic system. We are expected to learn, but are never taught how to learn. This is important since real-world problems do not come in the form of Math’s, History and Science, but can require a combination of all and being able to learn quickly can be a lucrative advantage.
“We know what to learn, but we are never taught how to learn”
We need Math skills, Programming Skills, Problem Solving skills, Business knowledge, Communication skills and other soft skills. These are all things that we are told to learn, in fact, we ought to learn them, eventually — I do not disagree. I do not dispute having these skills would not be necessary to progress in your career as a Data Scientist. But, when these requirements are pitched as “How to Become a Data Scientist”, it gives off the unrealistic illusion of a there being a destination that when arrived at will render someone to be qualified as a Data Scientist — Which is a complete misconception.
Being equipped with business knowledge, statistics and knowing how to program, for example, does not necessarily mean you can successfully apply these methods to unstructured or structured data to extract knowledge and insights, of which is the purpose of Data Science in the first place. Though it provides a competitive advantage, using Kaggle competitions as an example, we constantly see Math PhD holders, and domain experts are beaten by someone that is self-taught, or whom has transitioned from a completely unrelated discipline.
“Trade your cleverness for benevolence” — Jim Kwik
These people know that there are essential things that must be learnt that is there to improve the underlying ability to extract insightful knowledge and insights. If they don’t know what it is or how, they know they can always find out.
The fundamental skill of learning to learn (or meta learning as it’s also referred to) engulfs all of the subjects we ought to learn to become Data Scientist. We would have to learn them anyways, but if we learnt how to be better learners instead of just attempting to learn Linear algebra, for instance, when it is required of us to know Linear Algebra to solve a Data Science problem, we’d be much better equipped to face the challenge because we would know how to learn effectively so that we overcome that challenge.
However, what we tend to see, especially from beginner Data Scientist, is that when faced with a problem that does not fit in the category of subjects they have learnt, they are much more poised to avoid attempting to take on that challenge because it is in the unknown — this is the completely opposite for the top practitioners of Data Science, my favorites being Andrew Ng and Abhishek Thakur.
The field is constantly change and to be at the top, we must be continuously learning and there is no final destination. Techniques that won competitions in the past are not techniques winning competitions today and to stay on top, we must learn these methods, fast!
“Learning to learn quickly is an essential skill for the 21st Century” — Jim Kwik
Final Thoughts
Photo by JD Mason on Unsplash
In hopes of boosting my ability as a Data Scientist, I have been constantly listening to interviews with my favorite practitioners, reading blog post from established Data Science writers, reading books to find the essence of what makes someone realize their potential. It has been so amusing to identify a common trait in high achievers is their ability to learn a topic quickly. We need to be learners, it is essential in the 21st century.
As many people are deciding to become Data Scientist now, it is very important that we make the deliberate effort to learn how to learn if we want to embark on a long career in the field that is constantly changing, since this will never stop the moment you decide you want to become a Data Scientist.
On the other hand, one may argue that Data Science is an quite largely academic field and it could be argued that academics are assumed to be learners therefore in stating all the prerequisites, it is implied that the person wanting to become a Data Scientist is already a learner.
If there is anything that you think I have missed or some points you don’t agree with, your feedback is valuable. Send a response! If you’d like to get in contact with me, I am most active on LinkedIn and i’d love to connect with you also.