Data science in 2017 — Part III
In my earlier posts, we saw the demographics and work profiles of data professionals. Let us now see what is actually required for getting into data science, from the perspective of those already in the field.
What skills should one have?
We saw in my earlier post that Python is the most dominant tool used by data professionals today. Naturally, this is the highest recommended language to learn in the survey, along with R and SQL. These three can be considered the holy trinity of data science.
How can one acquire these skills?
There are a ton of online courses available on data science, most of which are free. Coursera, edX and Udacity offer excellent resources and active communities that help pick up Python and R skills, along with data visualization techniques. Udacity also offers career coaching, resume reviews and help with interviews. After learning these skills, one may head to Kaggle to start analyzing some real life datasets and competing against (and getting support from) the world’s best brains in the field.
How can one get a job in data science?
It’s all about who you know, I guess! Even though one may have all the skills required to be a data professional, networking will go a long way in finding a job. A majority of respondents said that they got their current job through referrals. Also, since data science is an evolving field, it’s important to keep learning. New methods, algorithms and applications of data science keep emerging every day and one needs to keep oneself abreast of the latest trends to be competitive in this field. And that’s why ‘learning’ was quoted as the most important factor in a job by the survey respondents.
What’s next for data science?
When asked about what they are most excited about for the next year, deep learning and TensorFlow were the top answers. I think 2018 will be when deep learning becomes mainstream, and more data professionals will resort to neural network algorithms leaving behind the more conventional methods that require a lot of preprocessing, feature selection and engineering. We may also see the emergence of reinforcement learning used in conjunction with deep learning. It’s also important to note that AWS and Google Cloud are featured in the top 10 answers. I’m surprised that Azure ML is not in this list.
The most important data point
At the end of the day, if you’re not going to be satisfied with your job, you shouldn’t take it. Data professionals reported a high level of job satisfaction in this survey. More than 50% of them rated their job satisfaction 7 or more on a scale of 10. Data science can be as rewarding as it can be challenging. Considering the lack of talent available, low entry barriers, tremendous learning opportunities and competitive salaries, data science may be the best place to be, if you have the right set of skills. Good luck!
