Matching and Dynamic Pricing in Ride-Hailing Platforms
HTML-код
- Опубликовано: 7 сен 2024
- Ride-hailing platforms like Uber, Lyft, Didi Chuxing, and Ola are transforming urban mobility by connecting riders with drivers via the sharing economy. These platforms have achieved explosive growth, in part by dramatically improving the efficiency of matching, and by calibrating the balance of supply and demand through dynamic pricing. We survey methods for matching and dynamic pricing in ride-hailing, and discuss machine learning and statistical approaches used to predict key inputs into those algorithms: demand, supply, and travel time in the road network.
The dynamic adjustment of prices ensures a reliable service for riders, and incentivizes drivers to provide rides at peak times and locations. Dynamic pricing is particularly important for ride-hailing, because pricing too low causes pickup ETAs to get very long, which reduces the efficiency of the platform and provides a poor experience for riders and drivers. Pricing and matching are intimately connected; we show that flexing wait time for riders during high-demand time periods can reduce the price variability caused by dynamic pricing.
See more at www.microsoft....
I keep dreaming that one day this channel will allow me to see the slides for more than two seconds at a time.
were they expecting her to build uber in 1 hour? what an audience. can't even count the number of times she said "this is a very simple model"
The audience is quite disastrous haha. Just wait for her to finish her presentation before asking
35:19 to 35:52 Platforms have to pay for the insurance when drivers are doing longer en-route distances to pickup (dead mileage), and the drivers pay more for gasoline and wear-and-tare on their vehicle.
49:23 I remember a passenger and I were once pondering the possibility of the platform using us as guinea pigs for extracting traffic data on an unusual route. I wonder how often such a thing might be done, if at all.
Informative talk.
Quite Informative :) Thanks for sharing.
Very informative. Thanks!
For all the questions that the audience asked, they were dead quiet on the "surge multiplier" concept (min. 56). Basically, she knew (judging by her look) that no one understood it. LOL
I don't think so. I left my opinion in a separate comment regarding min 56.
I had a feeling this problem got overcomplicated while listening to it.
Can someone share the link for the paper castillo wrote?
nice presentation, and way to much interruptions from the audience
Yeah, I think its better to accept questions at the end of the presentation.
where are slides
While the speaker did a great job giving a sense of their operations I felt some novices might find it hard to follow along. I found this video to be a rather little better at dwelling on the basics. ruclips.net/video/-EnghNB8dSE/видео.html
The flow of the speech could be better .But thanks for sharing.
I may not understand this well enough, but I feel like the presentation is quite disorganized, just presenting all the stuffs where she and her team replicate results from multiple sources of papers. I may be too picky because she is not an Economics PhD, but her use of economic terminology like equilibrium, efficiency, and welfare is so loose, that It's quite hard to understand what she is trying to deliver in this presentation. For example, she was not able to capture the intention of a question from a gentleman here(56:10). The concern that he has(I guess) is that the equilibrium concept is not even well-defined in the presentation, and there may be other equilibrium that this model may converge to. This is pretty critical one as the welfare implication may depend a lot on which equilibrium the model ends up producing if there exist multiple equilibria of the dynamic pricing model.
And I think the questions from audience are pretty fair, as many of them point out what's gonna happen if we take into account some other features that this model does not consider. The equilibrium result easily breaks down even if you add minor modifications, and this will result in totally different welfare implication(Not sure whether the model is pretty robust to this dimension).
Can you at least publish the slides?
lmao, the audience at the front ask the worst questions
Obnoxious audience
Im sure she is knowledgeable in her field but Uber Maps is much too far behind Google Maps. There really needs to be a change at Uber
Guys, I want to see the presentation rather than the presenter.
I mean they came to butcher her presentation. LOL, I think she was nervous, she knows the questions but the answers are not convincing enough.
Microsoft is one of the investors in Uber (or was). So...LOL
I felt bead for her, not having exact answers to the questions from audience. T_T I guess if there were other people with her, they might have helped with answering tho. Haha
I fellt the same too. She is not quite conversant with what she is presenting.
Uber has never been able to be profitable.