Basically new cases on Wed and Thu were higher than you would expect, even accounting for the fact that Wed/Thu normally see a higher-than-average number of new cases due to reporting effects, which indicates a changing trend.
Show this thread
I have this kind of crappy model for new cases that uses high frequency case data with a seasonal adjustment to predict where smoothed new cases will be in three days. Today is the first time for over a week that it predicts cases starting to rise (in the UK)pic.twitter.com/dYdudZSl7o
-
-
-
But also - it's clearly a v noisy prediction! Completely missed the shape of the recent peak for example. So this could just be noise and maybe cases will continue declining
Show this thread -
Also I am very very not an expert, just a nerd with easy access to data analysis tools. Thought it might be interesting to note given my convo with
@M_PaulMcNamara this morning.Show this thread
End of conversation
New conversation -
-
-
Sad times, but nice model :)
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I predicted deaths in Brazil really well for 3 straight months and then I got tired and stopped. Simple assumption: nobody knows what they are doing and everything will continue the same. Just used a linear regression in a loglog chart for support line.pic.twitter.com/eeuGmuepvo
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Not bad dude
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.