Okay, so blogging is hard. Well, the blogging part is easy but finding time to blog is difficult. Anyway, for the moment I’m going to ignore the fact that my last blog post was over six months ago in which I proclaimed I was back as a wannabe data scientist. Let me clarify- when I said I was back, it was more like the 2016 version of the University of Texas Longhorn football team being back. In the first game of that season, the Longhorns matched up against the No. 10 Notre Dame Fighting Irish. Texas won a thriller in double overtime, and as Tyrone Swoops dove for the winning touchdown, the play-by-play announcer said emphatically, “Texas is back, folks!” What we all didn’t know at the time was that both teams would end being really terrible, so saying Texas was back ended up being a bad joke for Longhorn fans and a favorite saying for Texas haters for the next two years.
(Oh man, I’m really digressing, aren’t I? I just hope it doesn’t take two years for me to be back.)
About the update…
First, we moved from Round Rock to Leander. We found a new house in a new neighborhood that’s five minutes from the girls’ school. Just about everything worked in our favor but I don’t want to move again. I mean, we received an offer above asking price after 6 hours on the market, we got the house we wanted, and the whole process took about two months. But the packing. I hate the packing. If I have any control over things, this will be the last time we move.
In the “I’m back” post (link here), I said I was looking into three machine learning competitions: Men’s NCAA Basketball March Madness, Microsoft Malware Prediction, and a reinforcement learning challenge. Here’s how they turned out:
- Google Cloud & NCAA® ML Competition 2019-Men’s - I expanded on the model I submitted in the 2018 competition and did well, placing 39th out 866. That’s quite an improvement from 2018, where I placed 498th out of 934.
- Microsoft Malware Prediction - I worked on this one a little, made a submission, then left it. If this one comes around again next year I’ll give it another shot. 1158th out of 2426.
- First TextWorld Problems: A Reinforcement and Language Learning Challenge - I played with this a couple of times but then ultimately abandoned it when we started the process to move.
So what’s next?
First, I need to publish my code for the March Madness Kaggle competition to Github and post a write-up. Second, I have a couple of ideas for Python web scraping projects that will grab some data and send me a text message. Third, I’m looking into a potential proof-of-concept project for my job built on Pyspark, Kubernetes, and Flask. There’s interest in a web application where users can request and submit Pyspark jobs to wrangle some data, run it through some machine learning models, and present the results in a visual format. Who knows when I will be able to work on this. As I said before, I’m only spending about 10% on these projects and my business analyst work takes a higher priority.
Definitely some things to keep me busy over the next couple of months. Busy enough to prevent me from blogging? Maybe.
Don’t mind me, I’m just rambling.