In my blog to date, I’ve taken a “tree approach” to explaining some data science applications. That is, I’ve talked quite a bit about the specific details of data science and tried to present them to a non-technical audience while not boring the technical audience. Today I want to start a four week series on the anatomy of a data scientist inspired by a FICO image I saw recently. I will talk about the analytical side of data science, explore the use of statistics in data science next week, talk about the role of machine learning, and finish by discussing the importance of leadership and communication skills.
A good data scientist approaches business problems analytically. They are naturally curious about connections and links between different sorts of data. Someone who is relentlessly curious is someone who keeps asking tough and seemingly strange questions when no one else does. It’s the unconventional out-of-the-box thoughts that are at the center of an analytical mindset.
Knowing which problems can be solved with the implementation of data is an important part of the analysis. I recently heard a talk at a Women in Data Science event where the speaker mentioned problems that have rules defined by humans make great data science problems to solve. For example, predicting traffic patterns with data works since there are rules we all agree on for how to drive in traffic. Curing cancer entirely might not fall within the scope of what problems data can solve since there are no rules (that we know of yet) determining who gets what type of cancer and when they get it.
According to Provost and Fawcett in their book Data Science for Business, when faced with a business problem, thinking “data-analytically” means having the ability to assess whether and how data can improve the process or product. They suggest that even if you never have intentions of becoming a data scientist, you will most likely interact with one and need to understand the context of their analytic mind set. Having this understanding will help the competitive advantage that comes with using data science to solve problems. Next week we’ll look at how the data scientist uses statistics as part of the analytical and problem- solving process.