Loved the presentation! Looks like you guys are having such much fun drinking beer and laughing around while presenting complex SOTA deep learning topics :D
Nice coverage of the paper. I have one question: you explain that the pseudo-label is used only if the confidence is above a certain threshold (which is also a hyperparameter). Where is the confidence coming from? It is well known that the confidence score coming out of softmax is not reliable. Can you please explain?
Hi everyone, I got some requests for the notebook source, I've cleaned it up a little and added a few comments: github.com/andykitchen/numpyo-examples/blob/master/sigmoid-changepoint.ipynb
Thanks for posting the video, I have missed the talk unfortunately. I have two questions: 1) Is the notebook available? 2) Poisson is a single parameter (mean) distribution. The higher the mean, the higher the variance. Is it correct that this will lead to a bias towards higher mean for regions with a higher spread of data points (earlier years)?
Thanks for watching! 1) Just put it up here: github.com/andykitchen/numpyo-examples/blob/master/sigmoid-changepoint.ipynb 2) Yes, this is a very insightful observation. Higher variance would be evidence of a higher mean and should push the posterior higher; this is the correct inference only under the assumption of the Poisson observations. If this assumption was strongly violated, it would cause problems, e.g. if the mining disasters where overdispersed (one plausible mechanism would be if a small number of very dangerous mines were opening and then getting quickly shutdown) During posterior predictive checking one would look for residuals much larger than would be expected. If you did find overdispersion, you could use a gamma-poisson/negative-binomial observation model or add per year random effects to the model.
The problem starts right after 5:15 and gets fixed right around 17:00. Audio and video are not in sync between these times. Most of Mat's talk is unaffected.
Thanks mate, the problem has been fixed in the source video, but I'll need to re-upload at a new URL. A fixed version will go up in the next few weeks.