Meeting with Prof. Devika Subramanian
My meeting with Professor Devika Subramanian at Rice was particularly interesting because she had done work very similar to mine. She worked on predicting conflicts in the Middle East, and also ran into issues with extremely small sample sizes when working with predictive modeling. Because of this she was both more sympathetic and critical to the issues I faced in my project on the Thucydides Trap. Since my project only has 16 data points, Professor Subramanian pointed out that I might need to do some adjustments for such a small sample size to even predict future events.
The main takeaways I had from our conversation was that while it is changing rapidly, machine learning and AI are still far from being able to compete with humans at certain tasks. One example is converting word problems to equations. I was also given some very good general advice on what to do when faced with tough decisions. One thing I found prevalent among most of my meetings was that I should learn Python, something I plan to do over the summer.