logs/week12.md

In wjones127/thesis: Bike Routes Thesis (Title Case)

Thesis Week 12

Models

I read Chapter 14 of the Gelman book and wrote up an outline of a few models. This can be found in ../thesis/models.pdf. Having written that, I think a good step to take soon would be to create an R script designed to answer essential questions about what kinds of models would be more appropriate.

Incorporating other data sources

I haven't gotten to incorporating other data sources, like road infrastructure data, yet. However, I did have some good conversations with my brother-in-law, who reminded me that we have invented data structures (kd-trees, to be specific) to handle problems like finding the nearest neighbor. So I haven't tested this, but I think I can make use of nearest neighbor search for some of the data mapping, which will be a lot more efficient than my previous attempts. The Fast Nearest Neighbors package will be useful for this.

Evaluating the schedule

According the my schedule from week 9, I was supposed to have all my data together and documented and start modeling. I'm still pretty close to where I want to be, and I think the most important thing I've done is the modelling, which has given me a better idea of what direction I will be heading.

So perhaps a new schedule?

Dec. 8: Draft introduction and write up a list of EDA questions for determining what kinds of models to run.

Dec. 15: Thesis table of contents, bibliography, introduction and data and data docs done.

wjones127/thesis documentation built on May 4, 2019, 7:34 a.m.