@paulgp if you haven't already, check out my work on issues with measurement in diffusion processes :) https://t.co/JI0MPb3dTx
2,485 followers
2,485 followers
My personal hobby horse here is to figure out a way to improve threshold estimation (and therefore cascade prediction). I find this problem fascinating. If you want to see where intervals come from, our paper is here https://t.co/4eCrwUkW6x
2,485 followers
@victorveitch @icristali1 You also may want to check out our paper on accurately recording exposure levels at adoption when using observational data. https://t.co/JI0MPb3dTx
2,485 followers
@ryanjgallag I def agree with this point but I wonder about practicality, because (as you say) networks are super complex. We spent a ton of time just to come to the realization that the *networked process itself* doesn't tell you node adoption thresholds