Not a technical degree- but very relevant for both Wall Street and Main Street-- just a good old fashioned major in Applied Math. Whether from a research university, a Liberal Arts college-- anything with academic rigor.
Someone who has worked with large datasets and is proficient with R, STATA, Matlab or similar- they don’t need to have a content-based technical skill to be a strong contributor right off the bat. I know “data science” sounds fancier than plain vanilla statistics… but vanilla works too, and is available at a much broader range of colleges! Kids slice the bologna WAY too thin…
5 Likes
I agree with this. I will give you one example.
I have seen multiple times in my career where we were using off-the-shelf software packages to solve specific problems, often involving some analysis of real-world data. In a few cases we were getting results that did not look right. What was necessary was to have someone who solidly understood the underlying math look carefully at the problem, and also look carefully at the off-the-shelf software, and figure out what math the software was using and what math was actually appropriate to the problem.
In multiple cases the software was doing something that seemed right to whomever wrote the software, but the math that the software was using did not fit the specific problem that we needed to solve.
This sort of misfit is something that is more common than you might expect. We were fortunate in that the results we were getting were so bad that we knew that something was up, and we also had someone on staff who was able to figure out what was wrong.
I have heard the same thing being done with machine learning. Again, someone who actually knows what the software is doing needs to be involved to figure out whether or not the off-the-shelf software is actually doing the right thing.
4 Likes