Why Become a Data Scientist?
Glassdoor has reported that the #1 job in 2016 and 2017 was Data Scientist. Springboard calculated there are 215,000 Data Scientist openings on Indeed.com. Simplilearn says the world will generate 50 times more data in 2020 than was generated in 2011. These statistics are why you want to consider becoming a data scientist.
Minimum Requirements to be a Data Scientist
You must have an interest and an affinity for working with numbers, an interest and an ability to code, very good communication skills and most importantly you are someone who is curious enough and clever enough to be creative in your approach to the numbers and the coding.
Since there aren’t that many candidates who meet the high expectations that have been set for Data Scientists there is still an opportunity to break into this role without perfect skills. The definition of a data scientist is not so fixed that there is an opportunity for the candidate with some of the skills or a history of getting/pursuing those skills.
What Do You Need to Do to Become a Data Scientist?
- Get a Master’s degree or attend a bootcamp. You must have advanced math or statistical skills and be able to code to call yourself a Data Scientist. In fact 63% of current data scientists have a Masters or a PhD. In a formal academic program you will get the training and the opportunity to do applied projects. Often a very good bootcamp or a strong online program can be as good as a traditional university Master’s degree.
- Become an expert in a language. With the prevalence of open source tools the ability to independently become an expert in R or Python is within your reach. Commercial products like SAS are harder to do without working at a company that buys this product, but not as important as it used to be since R, Python and SQL are the top languages for Data Scientists.
- Work with Data at the scale of the web. Find a way to work with data of the velocity, variety and volume of big data. Offer your services for free on a project. Create a project at work that allows you to do this. Find some way to use Hadoop, MapReduce and or Spark. Any exposure to any of these products will be invaluable for your career.
- Clean Data. Volunteer for the job every Data Scientist hates: cleaning the data. Get experience at the collecting, cleaning and validating of data sets for your firm. Sure data exploration and modeling is the fun stuff, but you will have real value if you can show your ability to merge legacy data and making sure large amounts of data are clean.
- Build a portfolio. Dan Saber, Data Science hiring manager at Coursera, advocates providing a portfolio of projects that show a range of skills to prospective hiring managers. Instead of trying to convince the employer of your skills, show them. Do it through a github or blog link. They can look at it at their leisure. Dan says, “Anybody who took time to put this together is interested. Anyone who could put this together is probably competent.” He goes on to say, “Hiring managers are looking more for evidence that you can learn new skills and apply them to real-life problems than they are for a laundry list of certifications.”
- Immerse yourself. Do a Kaggle competition. Do meetups and join online communities. Do a free project for a nonprofit. Convince your boss that the benefits of that Hadoop project will be bottom-line significant. If you are unemployed, do a free internship in the right type of analytics or coding.