Transcriptions
Note: this content has been automatically generated.
00:00:07
sure i'm going for a first it's u. e. r. h. b. o.
00:00:19
uh_huh well you mean we d. in a physics bayes the softer ah them uh
00:00:26
uh it's not really a choice i'm gonna do it but i still have to do it if that's the reason
00:00:34
'cause uh i think a lot of feelings means the tram very long time for um
00:00:41
put the simulation a running yoshi a scene was asking for three
00:00:47
months or so it was really a lot they had to divide the
00:00:51
um the city for tuning in a different is
00:00:55
more their ass and calculate every every yeah so
00:00:59
to the mid time consuming and uh considering both so the uh
00:01:08
uh this tuning was too much for for the first phase but i will do it
00:01:15
or or or or or does it yeah where we are for it yeah
00:01:29
yeah well when i actually uh and that uh that was one
00:01:33
to seven was the mean a performance improvement the mean every improvement uh
00:01:39
in some cases i had the first single building like twenty five percent of improvement i think it's a lot
00:01:46
especially building that had like a really a lot of fair or uh uh in the in the forecast
00:02:08
okay uh i don't know why it's not working but so one that's is uh it was a scatter plot between
00:02:14
the in measured uh and and the melody and the
00:02:18
measured consumption on the x. x. is and the t. v.
00:02:25
um calculated one uh in longer off a guiding the artificial intelligence
00:02:30
and the other for the uh his gun these model on the way
00:02:50
yeah well yeah okay
00:02:57
think well that's because um and there is a lot of
00:03:01
research about this so called the performance yep so we know that
00:03:06
the magnitude of these performance gap could be in that in those ranges it's normal when your with
00:03:12
this kind of simulations and um so it just compared to to the litter to we can say
00:03:28
yeah you're right
00:03:49
okay well i yeah well um
00:03:58
for some parking meters i use i've used to some research is the were made before mean a special
00:04:05
for switzerland and tuning we had a lot of research
00:04:08
is on the most common uh was diets or um
00:04:14
in addition rates so it's basically assumptions uh i agree with affected i should
00:04:20
do a good then out of the artificial intelligence 'cause we wanted assumptions free
00:04:27
but uh for now i was helping the
00:04:30
position of the model mm because especially d.
00:04:35
we some well no uh right assumptions uh so for now they are there but
00:04:42
that there is still a lot to do and a lot to improve uh it's a ah
00:04:49
complex field we can do a lot of things
00:04:52
we also have to factor whether
00:04:55
the open future are so low things it to them