Whats_next
Wow…
We are coming to the end. The semester just like the year has flown by. This was a great course.
As I reflect on best practices, I will make it a point to look for simpler solutions. More times than not I begin with the most difficult solution to the problem. Not sure why but it’s been known to happen. This can lead to frustration on my part and wasted time. I also plan to sketch out the code for projects (especially with shiny apps) to stay focused on the results.
My goal (for now) is to become a decision scientist at a financial institution or insurance company. Initially, I’m looking for a more structured organization so I can gain experience and have support from more senior associates. Ultimately, I would like to work for Google or Amazon. Yes, I’m thinking big….
During the upcoming months, I plan to read more to fully understand the relationship between data science and statistics. (Honestly, this is one reason I decided to start this program.) This course has helped me solidify that there is a big distinction between the two. The scope of statistics in data science seems
sadly limited to intro stuff – measures of center, visualization, etc. Statistician don’t use many (if any) of the algorithms commonly used in the data science field. Honestly, I still don’t see how hypothesis testing fits into data science. I understand the testing concept. There are millions of observations and we sample 100k. So, Same concept - evidence to support or not on a larger scale?? When it comes to sampling “big data” at what point is your sample too large? Another concept that seems foreign to data science is simulations. If you have access to millions of observations, why do you need to simulate any data?
Why couldn’t you just run the algorithm on another sample (group not tested before) from the group? (I saw simulations on a posting for position over the summer)
Finally, I plan to continue working on my MS in statistics and look for alternative career paths as a business analyst, data analyst, business intelligence role, etc. Honestly, I am not confident that most employers are looking for stats or even math majors – even if you have a masters or PhD to fill roles in the data science industry. When you look at many of those openings, it’s focused on hardcore computer programmers – python, pySpark, Hadoop, experience with big data (now I’ve seen a local postings for a data analyst asking for experience with millions of gigabytes), SQL, NoSQL etc.
Hmm…this entry sounds kind of sad. It’s not. Life would be boring if you didn’t have change and learn to adapt. So check back often - you might read that I am still working on my MS and completing a second undergraduate degree in computer science.
Have a great winter break!