Learning Projects 2017

Published on 12 Feb 2017 ⸱ 3 min read ⸱ 16 views

Every year I attempt to learn something new. I give myself an entire year to learn it and learn it right. Except, over the past couple years, I started off learning various topics but never followed through and dropped off mid way. Last years' project was to learn python and automating the boring stuff. Then there also was the violin and Spanish. These learning projects are not quite ambitious but they're an attempt to keep learning new things and having a side hustle outside of work (work can get mundane and monotonous). I had to also figure a way to follow my learning project through and not drop off mid way.

For this year, I want to learn two things 1. Machine Learning 2. Behavioural Economics. Both these topics are interlinked with what I aim to achieve at the end of the year. A Credit Score. Lending in India is a huge problem. Both access to it and repayment. Not so much Lending as much as it is a data problem. Giving money to individuals is easy, getting the money back - that's where the challenge lies. Part of the solution is a mortgage a.k.a. having some sort of security of equal or more value compared to the amount you borrow. The other part of the solution is creditworthiness. How do you segment individuals as good borrowers and not so good borrowers. Intentional vs Circumstantial defaulters - the term that lending experts use.

I believe this problem is solvable with data and behaviour analysis. Machine learning will be crucial from the data analysis part and Behavioural Economics will be crucial from the behaviour analysis part. No shit, Sherlock. Both subjects have far reaching use cases. While they both appeal to me from multiple use cases, I will try to limit them to the above focus area.

Will be learning Machine Learning from Andrew Ng's coursera course and this Stanford Open Course class. Will be learning Behvaioural Economics from this MIT Open Course class. Honestly, there's very little about Behavioural Economics that at least I could find online. Apparently it is a very controversial topic among those in the economics field as there's not much proof that it works.

Anyway, will give an update to this post by the end of the year to show my progress or stagnation. If you come across course work that you believe will be helpful, please do recommend. You can reach me via the comments section or via email. Look forward to your recommendations/feedback.

Onward.

Note: All blogs posts till 2022 were migrated to this platform (react+next+tailwind). While all efforts were made to migrate wihtout any loss, the migration lost some images and broke a bunch of links in old posts. If you spot anything amiss, please notify me?