Things starting to get interesting this week…

Week 03 Topics:

We covered more stats (t-tests & z-tests, Hypothesis Testing, Frequentist vs. Bayesian, A/B testing/Multi-Arm Bandits), Exploratory Data Analysis (EDA), and Linear Regression.

Of these I think the topics related to experimental design were the most important and interesting. The examples shown in class for A/B testing and Multi-Arm bandit strategies gave a really direct showcase for how important a strong statistical foundation is for making correct product decisions and not straight up playing yourself.

Plus this pictures is doubly hilarious when you understand what’s going on:

Multi-Arm Bandit

With the linear regression classes we began doing what I’d call “real data science”, the cool stuff.

The individual and pair exercises were fun since we were tasked with fitting LR models to our data and improving them. It’s pretty satisfying to see some crazy sh*t with your data & initial model, correctly interpret what’s wrong with it, and then be able to make the right adjustments to improve it substantially.

Next week we are supposed to learn more modeling techniques and do our first case study, where we’ll be working in teams to try to build the best model we can from start to finish.

Should be a whole lotta competitive fun!

Cohort 17/18 Capstone Presentations:

On Wednesday we were treated with a glimpse into our futures through watching the final project presentations of the cohorts before us.

Some projects were obviously better than others, but my overall takeaway was “Dam, Galvanize must be doing something very right here…”

I mean, if you consider that most of the presenting students came into the program with only a limited knowledge of data science topics, and just 3-months later can present such rigorous and detailed reports, thats pretty darn impressive.

One of the presentations that really WOWed me was by Carmen Lai, who reported on Github open source package trends. The analysis, visualizations, and interactive app she created set the standard high for how polished I want my own capstone presentation to be, plus it was NLP related.

Given that, I had no choice but to get in touch with her for project tips and suggestions!

Hopefully, the advice will help me equally amaze the cohort lucky enough to catch my presentation.

Getting Indexed on Google:

“You never know what you don’t know until serendipity knocks it right upside your head.”

After making this website, I thought that eventually it would naturally show up on Google as one of the top results when searching my name, just like my social media accounts and old blogs used to (..before I took them down because of how embarrassingly bad they were). But apparently, in this regard having a site created from scratch on your own domain is very different from one on a blogging platform.

When this site never showed up on Google I assumed it must be because I haven’t done much with it. It wasn’t until this week, during a career services exercise where a partner Google’s your name to simulate what a prospective employer can find out about you online, that I discovered the problem.

After Googling me, my partner found my LinkedIn, Twitter, and Github pages easily, saying I have a pretty decent web presence. I casually mentioned that I also had a website but for some reason it hasn’t shown up on Google searches yet, to which he replied “oh, is it indexed?”. This set off ringing bells in my head like “holy sh*t, I actually need to do something to get on Google???”, and lead to me looking up what indexing is then realizing how deep my ignorance was…

A quick register for Google Webmaster tools , sitemap creation, and couple other code alterations made and the site started showing up on Google results, albeit only on the 2nd page for now (maybe I should invest in an SEO/SEM book?).

Painful realizing I should have done this long ago. But, oh well, never too late to learn right?

Til next week.

=)