Guys, I bet you've expected reinforcement learning, deep learning, or another sophisticated stuffs . This is real good presentation from someone (who I've never met) wich has huge experience. Good productionize model should consider computational efforts (business-dependent) , complexity , intepretability among other variables. All my respect.
Guys, I bet you've expected reinforcement learning, deep learning, or another sophisticated stuffs . This is real good presentation from someone (who I've never met) wich has huge experience. Good productionize model should consider computational efforts (business-dependent) , complexity , intepretability among other variables. All my respect.
I expected this talk to be on how the real-time bidding price is determined. This is more about relevance or pctr modeling.
The talk is from the DSP point of view (the advertiser) it would have been nice if publisher's point of view was also discussed.
This is good presentation for understanding what RTB is. For beginners at least
She barely scratched the surface. There is a lot to talk about on bid strategy, bid optimization, budget spending pacing, and dynamic pricing, etc.
yep
Great talk. I don't think she got the engineering question at 18:09
I realize that too. I think the answer the guy was looking for is why logistic model faster than some other model during inference.