Sorry, Online Courses Won’t Make you a Data Scientist Unless You Didn’t Work On Projects
Data Science is not a single term; it covers a variety of topics and networks, such as the Internet of Things, Deep Learning, AI, etc. In simple terms, we can count data science as a complete blend of data inference, algorithm computation, analysis, and technology that helps in solving multifaceted business problems.
The reason why you should really stop signing up for more courses without applying what you already know
It was a weekend. I had just finished another online course on Data Science. I felt accomplished. Well, after “successfully completing” 5 different courses and receiving “certificates” for each of them, anybody would tend to be of the confidence that they are now real data scientists. I too was no different.
But online courses can be a funny thing. Most of them have glittering descriptions, a huge list of topics that they would cover, promises to make you adept at one or more skills and if we are lucky, we can also see a bunch of testimonies from other participants that are usually about how the course saved a participant from perpetual doom and made him or her an absolute champion in the field. But, what really most of us look for are those final certificates that we receive.
Those online, colourful documents bearing our names make the biggest difference to most of us. But one day, we sit in a room across our potential employer and see that most of them seem less impressed than we expected with our certificates. Some don’t even extend the courtesy of acknowledging the time spent on getting those certificates. They just get straight to the point and say, “These certificates are useless if you have not worked on any projects”. Now, it’s hard when somebody tells you that. Even more because our hopes on doing well at that interview and getting selected was primarily pinned on the online courses we had taken. It was our investment of money, time and effort. And to have our primary weapon disqualified even before the contest would weaken anybody’s confidence.
Trust me. I have been there as a second year undergraduate, one who carried 2 copies of a minimalist resume, 5 different certificates and a huge bundle of hope into 7 different rooms in one day, each with a different company at the college’s annual internship fest. I sang the same song of “I have certificates from these courses…” in the first 6. In the 6th company, I received the knockout punch that came in the verbal form of “Look. It’s great you finished these courses. But, you have not worked on anything. You don’t have a Github account. We don’t know your capabilities. So, we are sorry.”
Ouch! That hurt somewhere deep inside. I could feel that usual clogging of my wind pipe that prevents me from talking clearly, something that’s very characteristic of me when I am hit hard by emotion. But, it was clearly not their fault. They showed me the reality, one that I had always been skirting away from.
We all come across that one time in our lives when somebody picks up a mirror and places it in front of our faces. This was that time in my life.
Well, the choice is ours whether to shut our eyes or keep them open when this happens. I decided to keep them open and that’s probably what’s made all the difference. That’s what makes me write this article.
Why work on projects?
The reason we are all so keen on completing courses is because we tend to think of an online course as a positive abutment to our academic degree and one that would be favorably viewed by employers. As a college degree is mandatory in several organizations to even be considered for a job, we nurse the opinion that an online course would be equally valued and considered as “Extra Learning”. Well, it would be and there is little doubt on that.
However, given the amount of exposure the internet provides to everybody, an online course is available to anyone. Therefore, even if we complete a course, there is no particular edge that we have over others who have done the same. And companies compare candidates because that’s their only way to easily choose the candidates they feel will suit their needs. Competition is in the nature of everything we do. So, the only way to stay distinguished from others is to work on projects.
Another reason to work on projects is to learn. Online courses definitely teach us a great deal, but they are constrained by the very force that prevents teachers from giving all their expertise to students in the classroom — The Syllabus. Online courses have to be planned and anything that has to be planned will have a trade-off of not being able to address all possible aspects of the topic in discussion.
On the other hand, if we work on a project, every step we undertake will cause us to learn a new concept. The mistakes we make will be far more than what we make while following an online course. However, if we are willing to learn from these errors, the knowledge we assimilate will be rather plentiful and useful.
Making that transition into project-based learning in data science
After being cut down to size in that interview, I went home as a determined individual. I was determined to start working on projects and not just bank on my certificates. But, making a transition from our natural inclination to a new practice is probably the toughest thing to do.
I read a few articles on how to apply data science and work on projects. Then I had a few light conversations with my peers in the following days. A few conversations were very discouraging, to be frank. Not that my peers were bad conversationalists, some of them just seemed way smarter than I was. They seemed to be doing very impressive stuff. Some were making a drone and the only time I ever saw a drone was on television. Funnily, I wanted to work with them the moment I heard that they were making a drone. But, I didn’t want to ask. I felt I was not good enough to work on cool projects like that. Sigh! We all make mistakes…
The most confounding challenge, however, was that I had only been used to courses where my progress was determined by weekly assignments. The grading system was formalized and somebody was grading me. In a project, things were different. I had to self-evaluate. And, I was unable to do this. I was never able to determine whether I had done something well enough. I was unable to be my own evaluator.
Sometimes, we are just too easily compelled to hand over the reigns of our life in another hands. And the force that compels us to do is often our own inability to understand our strengths and weakness
I realized that I needed to prepare myself to be my best evaluator. This is what I did. I sat down and drafted my project idea and even set goals to be achieved, ridden with time limits. Honestly, I overshot every deadline but made sure I at least finished 80% of what I had decided to do.
The whiteboard hanging in my room was witness to my daily planning, my cheat days, the concepts I learnt, the concepts I tried to redefine and the block diagrams I always loved drawing. My first project was to analyze chocolate bar ratings. A pure EDA project, it provided me with my first experience at working on a self-decided project. I worked on it because I was interested in knowing more about the chocolate ratings across the world.
It’s not difficult for us to work on a project if we feel connected to it in some way or the other. So, maintaining this personal touch and being excited by the cause than the tools used is important to being able to stick to finishing a project.
If that statement about being excited by the cause seems a bit difficult to understand, here is another article, where I have clearly been excited by the cause!
But, why is project-based learning not easy?
Often, I have switched off in between lectures at college and picked up a sheet of paper and scribbled furiously about ideas that would have just struck me. A good deal of these ideas never saw implementation because they didn’t seem impressive enough to go through with. It felt as though these ideas would not help me become a cool data scientist. Wanting to be a cool data scientist was too ambitious, given the fact that I was a far cry away from even being a data scientist (I still am). Yet, the hysteria of wanting to work on cool projects kept me chained to the pole of illogic. I was unknowingly searching for my drone.
Even the chocolate analysis project was initially shelved by me for not sounding cool enough. Thank god I reconsidered. Most of us shelve projects because we feel it’s not good when compared to what somebody else has worked on. It’s a self-harming thought, to be frank. We fail to see that no two people X and Y share the same background. So, we can’t expect to be as good or better than somebody else at all times. All we can do is try. The result is really not in our hands. Those who work on complex projects probably know way more than us.
We can’t expect to make a ship without knowing the property of buoyancy. Trying to do so would just be plain stupid.
So, do we want to be stupid ? I assume not.
The Final Point
There is never a perfect way to work on academic projects. This is mainly because every student has a different approach to work on projects. Some do it for the grades, some do it to learn and some for both. Some see projects as a way to work in their comfort zones, while others view projects as ways to learn newer concepts. The combinations and approaches to working on academic projects are many. However, a few important points I have noticed during my undergraduate degree in CSE is what I will leave you with here.
- Projects are not meant to be demoted to the status of sheer comparisons between ideas. Every project has a learning point and it’s always in our hands to see not what the project does not have, but rather to see what it does have.
- Open discussion on ideas and topics is always healthy for our growth. But, it’s equally important to be receptive to good feedback.
- Self-monitoring our progress on projects is necessary because nobody else is going to monitor us in a real-world.
- Finally, rewarding ourselves for even small amounts of progress will ensure that we never lose our motivation
Finally, I don’t discredit online courses. Some certifications really matter to employers. Also, courses are a wonderful place to learn specific skills. Some of them are taken by highly reputed practitioners and equip us well. In fact, I learnt my first steps in data science from the wonderful Datacamp. But, the equipment alone never makes you good.
The above story is totally based on my friend Ramshankar Kumar experience in data science field.
Also Read :