Interesting application of clustering.

My group in the UK (Amey Strategic Consulting: https://www.amey.co.uk/amey-consulting/services/strategic-co... ) have been doing very similar things for a while now to optimise utilities maintenance, for example optimizing the number and position of maintenance depots to enable a utilities company to undertake repairs most efficiently.

I'm UK based, and not that familiar with the Bay Area - but I guess there are a whole bunch of follow-up questions. Most pertinently - what would this change. Suppose (for example) you created a North Bay Transit Authority - how would a common ticketing policy affect commuting patterns. Is the issue the cost and inconvenience of multiple tickets, or the disjoint nature of current services?

In germany, there are a lot of the "Verkehrsverbünde" and they work very well. For example, in munich there's the MVV (Münchner Verkehrsverbund), but the s-Bahn (slimiliar to BART) is operated by the Deutsche Bahn and the underground, tram and busses are operated by the MVG (the communal transit company). There are, of course, problems in most of the public-transportation networks in germany, but I don't think they are to blame. I think it's a pattern worth copying.

It even allows for some competition because you separate the provider from the system. For example, in nuremberg the provider for the s-bahn for next few years was recently auctioned off to another company after frustration with the deutsche bahn.

This is cool. I'm working on something similar to divide up areas for municipal waste collection based on travel times between properties.

> for this problem, we are defining the distance matrix directly from the source data. This results in the unusual property that the feature space distance from A to B is likely to be different from the feature space distance from B to A, as more commuters will commute in one direction than the other. For this project, we made the decision to examine origin to destination commute flows only, as this resulted in the clusters that were clearly defined in both feature space and real space, while the inverse resulted in clusters that were significantly overlapping in real space.

I had a similar problem, since the travel time from A to B can differ dramatically to that from B to A. I experimented with a few different ways of symmetrising the matrix and found that taking the maximum of both values was a pretty good compromise.

I also found that hierarchical clustering didn't work as well as K-means/medoids when the "clusters" were not necessarily very well-defined and the number of data-points was in the thousands.

This article was very underwhelming. It’s viz porn dressed up as something insightful. Looking at his clusters, he managed to “find” the transit systems. That’s not particularly interesting. It’s even less interesting because at the start he talks about how you can’t combine transit districts, and then he clusters to find the transit districts and then tries to say “look at all the transit districts I combined”, when in reality the ones he’s combining are tiny and inconsequential.

If he wanted to do something, he should demystify clipper card purchases. Every agency has its own bespoke fare system, and its complete bullshit if you transfer between systems. THAT would have been interesting, especially given how these things interact with agency funding.

Apologies for playing devil's advocate but in areas where there are few riders, would it not be a reflection of poor transit service in the first place? I would assume if the service is so bad in those areas that ridership would be poor but you'd still need to check out whether the actual residents in those areas are using alternative methods of transportation as opposed to just commute data?

Absolutely amazing research!

But let's now imagine the transit systems of SF area have been reshaped according to the article. It will surely trigger a change in commute patterns. Will this change be sufficient to significantly affect the clustering? Will the system stabilize eventually?

Isn't this in fact k-means clustering rather than "machine learning"?

This does not take into account potential riders, am I correct? This seems to be a common pitfall when planning new routes or systems.