Kyle McKiou is someone that aspiring data scientists should follow. He often posts ideas or materials that are helpful. I find his perspective refreshing. He gives sound advice related to learning data science, some of which I’ll share with you below. But most importantly, he’s a Gator! At least LinkedIn tells me he is, I don’t actually know Kyle.
(Also, he used to be able to deadlift 600+ lbs; that is a lot)
The first post from Kyle that caught my eye was this one. He mentions using Kaggle competitions to showcase your talents to potential employers. This is a great idea that I mentioned in my guide to getting a job in analytics in 2018 (get it here). He goes into more detail than I did.
The second post that I noticed was this one about math and business sense. It is no secret that many corporate data science initiatives do not create much value. Some times it has to do with the math. There are some very easy options for getting started with machine learning. You can upload data the data and have a “model” minutes later. However, that “model” will be useless or wrong without understanding the related math and data needs.
In my experience the issue is more often with the business side. (See my article / video on the five types of analytics jobs for more) The underlying business fundamentals are as important as the math. A helpful exercise is to lay out the precise actions someone will take with your results. If that is not clear, keep working with your business partners. You need to understand what is, and is not, actionable.
His final post that I want to call your attention to is about how to get started. I love how he proposes to get started. Find a problem that interests you, or you need to solve, and figure out what you need to solve it. Necessity is the mother of invention! The best techniques to learn are the ones that will benefit you. Identifying problems you need to solve and then the methods to do so is the best way to ensure they are relevant.
I wrote in the guide to getting a job in analytics that adding analysis to your current role was the best way to get started. He has come up with another reason why that is the best path. I had never thought of this, but it is perfect! I will incorporate this into the updated versions.
He has some great articles on LinkedIn here as well: