In response to my summer study plan, my adviser indicated that my search is hopelessly broad. I don't disagree that actually mastering all those areas would be way too much, but I don't think there's any harm in giving them all a cursory glance to see if any relevant nuggets can be mined. I think the problem is that I haven't adequately narrowed the actual area of research.
I was thinking about this on my run yesterday (thinking about math during a long run in 95-degree weather probably ranks alongside being waterboarded for most people, but I rather like it). I'm going to add a couple more adjectives to my final deliverable: it's now a MPP Framework for Multi-dimensional Analysis of Seriatim Financial Data.
Those two descriptors buy me quite a lot. First and foremost, by specifically targeting financial data, I can make a strong case for assuming the data come from Stable Distributions. This, in turn, justifies all the stratification in CISS to force convergence of sample means. It also opens up some avenues for pursuing actual mathematical results since Stable Distributions are a fairly active topic in their own right.
Secondly, the assumption of time series data opens up all sorts of data reduction techniques. My adviser has been pushing me to look at this for quite some time. I haven't been intentionally dragging my feet; I just haven't gotten to it.
No comments:
Post a Comment