Shameless plug - this was my own work (done as a 20% time project) kindly submitted by my wife.
Since it was a pro-bono project, it was hard to make it perfect but its a work in progress and i'll continue to make it better. I'd love any feedback so I can make it better.
There are numerous directions I can take the project, and since its not for a paying client, I have a lot of leeway to shift directions based on community feedback. Feel free to leave comments here or DM me if you wish.
Specific areas for feedback:
- Are commercial flights the best "network" to use? Alternatives?
- Are there good studies on the most impactful rubric for donations (deaths? cases? poverty? population density?)
Its an interesting math problem that i'd like to turn into something more real with feedback.
Vaccine supply issues are hugely complicated and one could model numerous things, including last mile. I focused on the long-haul optimization, but could have focused on last-mile as instead (or as well.) One could also focus on the hand-offs, delays, capacities along the way, etc. I can probably do all these, but curious what the core problems are -- from a total outsider's perspective.
I'm not an expert at graph optimization, so the entire effort was a way to force myself to learn some of the approaches and setups possible.
Most of the issues are bureaucracy and political, but I ignore that aspect of course.
Yeah, I don’t know anything, but it seems like understanding the problem well is important, and writing up whatever you learn about your inputs and assumptions (problem formulation) would be valuable.
Funny thing is that as we spoke internally about this project (20% time side project), there was a lot of discussion about whether to pick a real problem (vaccine delivery) or a totally fun one (minimize Ubers for a city-wide pub crawl).
The resolution was -- even if the objective was just to learn - we should focus on an important problem rather than a fictional/meaningless one.
That said, youre right, I think good formulation and assumptions would be critical here. This is the type of feedback I really want - hope to do a slew of commits on Friday.
Another possibility would be to lightly fictionalize it, so it’s obviously based on COVID supply issues (and that adds interest) but you don’t need to try too hard to be realistic and you can make stuff up wherever necessary.
This would be similar to what people do for games where playability is more important than realism, but you want an interesting theme.
TuringNYC 2021-08-18 14:43:18 +0000 UTC [ - ]
Since it was a pro-bono project, it was hard to make it perfect but its a work in progress and i'll continue to make it better. I'd love any feedback so I can make it better.
There are numerous directions I can take the project, and since its not for a paying client, I have a lot of leeway to shift directions based on community feedback. Feel free to leave comments here or DM me if you wish.
Specific areas for feedback:
- Are commercial flights the best "network" to use? Alternatives?
- Are there good studies on the most impactful rubric for donations (deaths? cases? poverty? population density?)
skybrian 2021-08-18 16:10:01 +0000 UTC [ - ]
If not, I would suggest rewriting the problem to be more obviously fictional.
Or if you do actually know something about it, tell us about that? It seems more interesting than the math problem.
TuringNYC 2021-08-18 16:15:02 +0000 UTC [ - ]
Vaccine supply issues are hugely complicated and one could model numerous things, including last mile. I focused on the long-haul optimization, but could have focused on last-mile as instead (or as well.) One could also focus on the hand-offs, delays, capacities along the way, etc. I can probably do all these, but curious what the core problems are -- from a total outsider's perspective.
I'm not an expert at graph optimization, so the entire effort was a way to force myself to learn some of the approaches and setups possible.
Most of the issues are bureaucracy and political, but I ignore that aspect of course.
skybrian 2021-08-18 16:24:11 +0000 UTC [ - ]
TuringNYC 2021-08-18 16:36:02 +0000 UTC [ - ]
The resolution was -- even if the objective was just to learn - we should focus on an important problem rather than a fictional/meaningless one.
That said, youre right, I think good formulation and assumptions would be critical here. This is the type of feedback I really want - hope to do a slew of commits on Friday.
skybrian 2021-08-18 16:46:37 +0000 UTC [ - ]
This would be similar to what people do for games where playability is more important than realism, but you want an interesting theme.