logs/week1.md

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

Week 1 Sept 17 - 24

This wasn't a super productive week for my thesis. But I have done a few important things.

Getting the Data

William confirmed that I can use the data, and we are getting coffee next Tuesday to talk more about the data and I'm assuming I will get it soon after that. In the mean time, I at least can look at what has been done in Rennie's project.

Started Research

Linda and I had a little trouble finding appropriate papers related to this topic. We did find one that was promising, in which they were creating a traffic model based on traces from taxi GPS. I think there are things I can learn from this paper, though I need more time to study it.

There are some smaller areas I think I will need to examine, such as best practices when analyzing data from online ratings.

Other thoughts

One possible way to look at paths is as a 'bag of streets', in the same way that text analysis sometimes model sentences with a 'bag of words' assumption. Then by splitting the GPS coordinates into observations of going through a street, I could take inspiration for a model from those text anaylsis models.

Plans for next week

I will explore the research done by Rennie on visualization, which will give me some good questions to ask William when I meet with him on Tuesday.

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