What it Takes to be Successful in the World of Data Science
By Saheli Roy Chowdhuri
Data science has evolved as one of the most lucrative careers in the tech domain within a span of just five years, owing to its universal application in driving business growth across verticals and sectors. Naturally then, a notable majority of youngsters today find themselves drawn to this up and coming career choice. To help such aspirants understand the dynamics of being a data scientist, Manipal ProLearn hosted a webinar on the different attributes of being a profession in the field data science.
The webinar, “What it takes to be successful in the world of Data Science?”, was presented by Gunjan Narulkar, a mainframe developer turned data science, professional, currently working as the principal data scientist at Fidelity Investments.
Understanding the dynamics of data science
Mr. Narulkar began the webinar by asking the attendees about their respective fields of work and relevant experience in order to get a sense of how well they understand the concept of data science and how they relate to it. He then tried to establish the fact that the growing demand for data science in the market is a reality and not just hype. If one understands the job environment, opportunities and the ground realities of data analytics, there can be no hindrance in riding this wave of opportunity.
Building blocks of data science
Somewhere in the 1920s, someone wrote a book where they described certain machines with human-like intelligence and the concept of Artificial Intelligence (AI) took form. Though the jargon has existed for decades and long before we had the means or technology to even venture into AI, the reality remains that even today we are far from creating machines that can mimic human-like intelligence.
However, this concept has set the tech world rolling in the direction of a larger goal. AI, machine learning, and data science are sets of tools that can help humans achieve that goal.
In short, data science is all about developing tools and then using these tools in tandem with statistics to solve different problems at hand.
Applications of data science
It has been established without doubt that there is a growing demand for data science in every sector and field of work today. The question to ask is: why are companies investing in DS? Well, because the outcome of data science aligns well with the end goal of every business – increasing their top line. The two variables governing this outcome are increased revenue and decreased costs for the sake of maximising profits. As a data scientist, a professional must focus on finding solutions to problems that tackle either one or both variables.
According to a Peer Research Big Data Analytics Survey, factors determining organisations’ willingness to invest in data science analytics can be divided up as 49 per cent decision making, 16 per cent better enablement of key strategic initiatives, 10 per cent better relations with customers, 9 per cent better sense of risk, and 9 per cent better financial performance.
It becomes pertinent then to view data science as a set of tools responsible for enabling a decrease in costs and an increase in revenues by improving the ability to make data-driven decisions rather than gut-based decisions.
Skills required to become a successful data scientist
Mathematics and statistics, as well as programming and database, are the basics on which data science rests. Of course, to be able to function as a professional in this field, you need to have a firm grip on machine learning, statistical modelling, experimental design, computer science fundamentals, programming languages such as SQL, Python, and R, supervised learning and reinforced learning, etc. You need to be adept at programming languages such as Python or R, Scala, as well as data processing frameworks such as Spark to have the collection of skill sets required to make you employable.
However, developing a skill set that suits your employer and the demands of the market, and at the same time has the capability of solving problems is equally imperative. Being successful as a data scientist requires a whole set of other tools and techniques as well. To understand these tools, one needs to understand the basic steps for problem-solving with data science analytics:
• Framing the problem which requires domain knowledge, product intuition, business strategy, and teamwork.
• Collecting raw data that entails database management, querying structured databases, retrieving the unstructured database, and distributed storage.
• Processing data for which you need scripting language, data wrangling, and cleaning, distributed processing.
• Exploring data with scientific computing, inferential statistics, and experimental design.
• In-depth analysis that requires machine learning and advanced mathematics skills.
• Communicating results that rely on business acumen, data visualisation tools, and storytelling, etc.
To able to work your way through this complete lifecycle of problem-solving using data science analytics, a professional must possess the following critical yet underrated traits:
• Commitment: Collaborating with a large number of people on a project, hearing them out and being heard in a scenario where everyone is an expert in their field requires commitment and a go-getter attitude.
• Creativity: Comes in handy when solving a problem where data is scattered over different sources. Relying solely on an algorithm for such a task can be immensely challenging. So, a data scientist needs to think creatively and out-of-the-box.
• Intuition: You may need to rely on intuition for tasks you haven’t handled before. However, intuition does not come overnight. It is basically taking an informed risk and for that, you need to have extensive knowledge of your field.
• Business Suaveness: It is a skill that cannot be taught and has to be acquired. The key to acquiring it is to align more closely with the business you are working for. It is only by bringing the context of business into your thought process that you can identify problems and find solutions.
• Presentation Skills: Numbers and statistics may make sense to the trained eye of a data scientist but these dry stats are of no consequence to a layperson. Unless you can weave a story out of your findings and put your point across in a manner that it resonates with your audience all your other skills are redundant.
These are really important attributes of a data scientist that go well beyond the purview of mathematics and algorithms and play an important role in helping you do your job well.
The way forward
As mentioned before, organisations look toward data science to find solutions that help them increase revenues and decrease costs. The innovation that has taken place over the past five years has shifted the focus of data science analytics from artificial intelligence to deep learning.
Going forward, data science will rely on a combination of data, advanced mathematical solutions to make sense of that data and technology that can deliver quick results so that the outcomes can be used in real time.
In this fast-changing scenario, a structured course in data science can not only give you a strong foundation but also help you hone your skills to be industry ready.