A little over a year ago, I transferred from working on applied mathematics in academia to working in the industry as a data analyst. Like any person that leaves the academic bubble, the outside world is very different from expectations. Working on data analysis in academia and industry both share common goals. They both aim to solve problems. But they emphasize different parts of the process. In my experience, academia focuses more on methods. They aim to test new methods and delve into unexplored areas of data analysis. The industry environment focuses more on the solution than the method. This means that the fancy algorithms and methods we learned in academia are not always the best option. Having an algorithm that converges to a more accurate solution, but takes more time or more computational power, might not be ideal.
I started to run into more and more problems in the industry that have solutions sometimes too obvious to be found. On the flip side, I have occasionally run into problems that require some intense statistical/mathematical/data analysis to get a handle on. The idea for this blog is to introduce data analysis for the industry, hopefully helping data analysts (myself included) come to better conclusions.
It is from data that we make conclusions. We do this every day. We predict how long our commute home from work will be at a certain hour, pizza restaurants estimate the delivery time of our favorite pizza, weather stations come up with forecasts, etc. Defining, harnessing, scrutinizing, and improving this prediction power will benefit everyone.
Enough idealism. The plan is to cover a wide variety of mathematical, statistical, and data analysis topics and show examples and how to implement them. The software that I am most comfortable with is R, Matlab, Python, and (gasp!) excel. If you want me to cover a specific topic or have a question about something, feel free to email me or leave a comment.