What the Hell is DIMMiN?

That’s a great question, I’m glad you asked and gladder you’re here! Click here if you want a broader overview of what this blog is all about.

A couple years ago I graduated with a BA in physics and accepted my degree in my underwear. If you told me that at the start of college I would have laughed and smiled, knowing that I had picked a really good school. I had snorted enough caffeine pills to get through undergrad and bubbled in enough right answers to get into an excellent data science master’s program. The plan after college was to complete my data science master's program, then start working as a data scientist in 2022, then sleep on big, big piles of money. Unfortunately celebration and long-term planning didn't match the general aesthetic of 2020. In other words, that’s not how things worked out for me (we will definitely circle back to the moneybed idea).

Things took a turn when I started the summer bridging course. My data science master’s program required me to take a 6-week summer course in data structures. This summer bridging course consisted of 3-hour online lectures MWF and cost me around $3,500 (even after a 50% scholarship). Typically I wouldn’t consider the cost too thoroughly and just take out a loan to pay for it, being assured that I would be rolling in dough before long. This course was different.

What really revved up the anxiety about having to pay so much for this course though was that it was moved entirely online. The professor was clearly brilliant and made the most of it, but I didn’t get to know virtually any of my peers (in a figurative way, in a literal way I know them all virtually). The entire experience felt pretty isolating. The real kicker was that I had just finished a data structures course in my last semester of undergrad so I already understood the majority of the concepts we were learning. 

Tuition for the remainder of the 3 semester program would be an additional $60,000. After loading up on $27,000 in undergraduate student loans, did I really want to commit to starting off my career with almost $100k in debt? Was I even going to be able to secure a job after getting my masters? Would I need to sell my kidney on the black market so that MOHELA didn’t break my legs?? Would our lord and savior Dave Ramsey part the clouds in the sky, descend from the great beyond, and smite me for my fiscal irresponsibility??? Probably not. But the question remained: in this economy, could I afford not to go to grad school? I began to wonder if there were other options.

Figure 1: Several different forms of collateral for student loan payments. Source: Robina Weermeijer

Just before the end of my summer course I came across a Ken Jee video which recommended using "Ultralearning" to learn Data Science. Ultralearning is a term coined by author Scott Young in his book by the same name. It is an incredibly efficient method of learning new skills and ideas. I flew through Scott Young’s wise words at mach speed, completing all 250 pages in under a month. After finishing, I decided to dedicate at least 90 minutes a day to teaching myself data science until I was hired. And that’s exactly what I did. Every single day for 366 days (except for three horrible, vile, wretched blips, but that’s a story for another blog post).

If I’m being honest, I didn’t turn into a data science deity in 366 days and wasn’t able to land my first data science job all on my own. I enrolled in a less conventional data science mentorship via SharpestMinds (highly recommended, they have a keen business model that solves the key issues of this higher-cost higher-education dilemma). After 471 applications and 102 cold-calls, I was hired and my life’s ultimate goal was fulfilled. I can now die a happy man and maybe even retire before that.

While I have accomplished my original goal, it has become very clear that I still don’t know jack about data science. Short story long that’s why this blog is here, for me to display the projects I create and discus the things I learn. I would encourage you to leave likes, comments, and other general feedback about what you think (as soon as I learn how to implement these features…).

After all, what would a data scientist be without your input?