logs/week20.md

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

Thesis Log: Week 20

Worked on chapter 4 for my thesis writing. It took longer than expected, mainly because it was a busy week. I have a pretty decent draft of most of the model specifications, but I need to do a little more research into smoothing splines so I can explain what I am doing with the non-parametric time terms and additive models.

Some questions for Andrew

• How do orals work? Should I be finding people for orals? What should I do to prepare?

2 people from math dept: Andrew and Albyn

2 people from elsewhere: Math and Natural Sciences Noel N. from econ, works on city infrastructure

Tuesday, May 3, 1pm in Library 389

• Should I include a time of day variable?

A new timeline

Ideally I could run a iteration of the first models in the next week, so that I could write those results up over break. To get those models ready, I need to

• make probability plots minimize MSE or AIC
• write function for additive model predictions
• write specification of each model
• implement cross validation
• test formatting of plots
• schedule model running time with Evan

I will make missing data models chapter 5, and write that up.

I will also start looking back at the route data and trying to work on that problem again. Hopefully I will come up with something.

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