Data Science Immersive [Week 4 Review]

main-qimg-2a88e2acc009fa4de3edeb51e683ca02.png

Oh boy, what a journey it has been so far.

One month into the data science immersive, I look back at the number of things I have learnt and I am amazed that I have managed to stay sane. From having almost zero coding experience to learning about multiple linear regressions (as above), I would rate the course a 9/10 for giving us a good taste of what data science is all about. That remaining 1/10 is reserved for my opinion that we should have less beer and more wine during our Friday drinks (sorry!). But yes, we do have beers provided by GA every Friday after class to celebrate the end of the week! #GAperks

Week 4 saw us learning about regression models, the introduction of scikit, feature scaling, cross validation and regularisation. There was also a session on object oriented programming, which I felt deserved more time than was allocated given how important the topic was. The week culminated in our individual capstone project pitches to the class, not forgetting the company open house at walkabout.sg where I visited the offices of Nugit, Propertyguru and Accenture. Accenture’s office was extremely impressive – understandably so since it was a showcase, but it totally changed my perception of them as being a traditional consulting firm.

image
Large spaces! Apparently the layout of the office on that level is flexible and can be modified when required
image_1
Accenture’s maker studio, which has two 3D printers and a robot bartender that dispenses mocktails!

Capstone Project: NLP

As my course mates and I took turns to pitch our capstone project, I saw how varied the number of topics were, in turn a reflection of our diverse interests across different industries. We had to name two ideas each and there was hardly any overlap in our topics, even for a class size of 25, affirming my view of the deep potential for data science to benefit businesses and individuals no matter where they come from.

For my capstone, I have chosen to do a variant of the spam mail filtering project where I will have to determine whether a particular news article is clickbait or not, using NLP (natural language processing). I have been told that it is not easy to do as the accuracy scores may come out less than ideal, but decided to go ahead anyway since I find the topic quite interesting. The next few weeks will see a slew of deadlines and projects coming up in addition to the reading/experimenting I have to do for the capstone. Still holding my chin up high in anticipation of greater things to come!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s