Bestand wählen
Merken

Trafficwerks [DEMO #11]

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Beta
Erkannte Entitäten
Sprachtranskript
no no no no no no no no you don't have move to the park so this
is traffic works I'm shot assist harmony and we're interested in the
climate change so we see everyone just got these flash-flood warnings rights but so am I going to say 1 flash-flood does not mean climate change on but we are getting a lot of weird rather recently on go the next so this is where I used to live on Austin Texas on a lot about flooding flooding in Houston as well I have a friend that just had supposed to pictures on Facebook that his whole house which is down and there's like hundreds of people are out and this this again this 1 anomaly does not show climate change seeing all these flooding and the West and the East Coast and receive drought in the west coast and so what can we do about this well we know that greenhouse gasses are somewhat to do with climate change and this 1 that 1 major contributor of greenhouse gasses gasses transportation and not a lot we know that it's probably about 25 % of New York on all the transportation of all the greenhouse gasses with anything we can do to decrease the the the how carbon is going into the atmosphere is something that we can be we can be proud of so what we want to do is take a transportation information from NYC DOT and bring into new pick and see can we
predict what's going on the traffic and if we can predict and see anomalies in traffic system then maybe we can re rout traffic to go around were places where there's on stops if traffic is blocked up then that makes cars idling having more on greenhouse gas emissions so cool after the pretty decent job this right now if you go to google maps you can see you right now that there's red lines place and that means things are blocked but we which is great except that of rush-hour always blocked off and so it doesn't really give you a lot of information to know that things are blocked right now or rather it's better to know what is really different from normal and that we think you can do we can see over the long period of time when when traffic is worse than normal and then and only then give notification saying she goes so that's the idea we didn't get far but we do we did get some information but so we went to the
NYC DOT and
went to the stream of and you nice it'd be nice if they had a nice API that we could pull data from or what they actually have is a CSV file which would be nice if this was actually queriable but this is real time data so any is no historical data so what we've been doing to get data is for the past where when and when did you say just start grabbing cached data right now yesterday morning been caching data yesterday morning every minute pulling all this data statues system and feeding it into new but that's the anything else about the data to the
values the data and we've got lots of around different paths so these black lines represent the pads that the traffic on speed cameras in New York City show odds all those black lines so all that data into new and we need people on just that all that little section right there that lit up and that thankfully and and I don't wanna seen me but that there seems to have been an anomaly so we did get to see that uh maybe get to see how well you pick learned about when that here OK so yes I can take that long so how will we do have a fairer we we don't have so the data comes in permanent so we have a fair number of data points we have just over we have about 560 i data points so when time runs this thing we're going to see is that and and this is really just from 1 of those lines we actually got data for about 200 different routes but this is just 1 of them we can a cherry picked 1 that had an anomaly so we see here going up and down the and again this is really just for a very small period of time and just during the week so what we would expect is we get this data for weeks at a time and then we would see traffic going up and down but right now we see red lights at 2 o'clock so then our our cash account crashed and so that's in this whole life is it ran out of memory on so we came in the market in the next morning and said all that's that that's the right to have of 2 to the coming up with OK so others so that we got now we go and so
on down at the bottom you take a word for it is the the prediction what is it the um anomaly so right about there we got to about 500 data points and then it said hey we could see and right here you see this huge jump right here which is it might be just around when the weather starts right and here we see this flat and just until we get about 5 and data points and now we can actually start predicting things that so what we'd like to do on given that the data is still kind of weird we can continue to run this thing put up to her room to instance and just start keep keep going going going most of the code is there but in fact the code light up a little line and Google Maps is they're mostly on and I think that it would be interesting to see this sort of thing after about a month in silicon readout and but just as this as an interesting aside this this this sort of prediction didn't actually take all that much processing power so we were actually able to run on 120 instances side by side but before we started slowing down and screwing around with the actual speed at where we can access it so instead of graphing 200 120 lines which is what the real life used to detect the anomaly right and that's it thank you do a frequently do you get the data I mean every minute so that the data file is every minute and were grabbing at every minute when is changed dramatically remittances reversal genuine up yet going you can see from the from the blue line that it's actually pretty consistent and where and and sometimes like during rush hour it'll change drastically that it'll it'll go from like we we we graph travel time and that its speed because we want something it would go up if there is something that would go down traffic and that will also so keep in mind that this is just 1 of the patents on which is fairly flat the first one we did it on it was more of a gradual up and down so it really depends on behalf the of the well and again there's always the risk of over sampling of the changes drove slowly compared sampling rate and then sort of work well but you sort of happy medium what we would expect if there was a really bad like that it would go up and up and up and then it would go from maybe 20 to boom and all the things we really want to predict like this thing here but it might actually help to have other types of data sets in the prediction algorithm are going to take into account the weather or if there is a concert or something else that we didn't quite have the time to implement those I but in the long run that's the kind of thing that we're going try to implement in the in the basic idea of not modeling traffic but the anomalies in traffic flow that's really cool is that banks right things again
the at the end of of the term the and and in the
Quelle <Physik>
Computeranimation
Demo <Programm>
DoS-Attacke
Arithmetisches Mittel
Facebook
Skalarprodukt
Rechter Winkel
Notepad-Computer
Vorlesung/Konferenz
Information
Transportproblem
Skalarprodukt
Rechter Winkel
Prozess <Informatik>
Vorlesung/Konferenz
Routing
Information
Physikalisches System
Frequenz
Gerade
Programmfehler
Videospiel
Punkt
Echtzeitsystem
Rechter Winkel
Zahlenbereich
Garbentheorie
Routing
Physikalisches System
Elektronische Publikation
Frequenz
Brennen <Datenverarbeitung>
Gerade
Graphiktablett
Prozess <Physik>
Punkt
Mathematisierung
Term
Code
Computeranimation
Wechselsprung
Informationsmodellierung
Algorithmus
Prognoseverfahren
Reverse Engineering
Datentyp
Stichprobenumfang
Minimum
Vorlesung/Konferenz
Gerade
Demo <Programm>
Leistung <Physik>
Videospiel
Graph
Bitrate
Elektronische Publikation
Datenfluss
Quick-Sort
Programmfehler
Quelle <Physik>
Menge
Rechter Winkel
Wort <Informatik>
Instantiierung

Metadaten

Formale Metadaten

Titel Trafficwerks [DEMO #11]
Serientitel 2015 Spring NuPIC Hackathon
Anzahl der Teile 19
Autor Carswell, Thomas
Lauzon, Shawn
Lizenz CC-Namensnennung 3.0 Unported:
Sie dürfen das Werk bzw. den Inhalt zu jedem legalen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen.
DOI 10.5446/18065
Herausgeber Numenta Platform for Intelligent Computing (NuPIC)
Erscheinungsjahr 2015
Sprache Englisch

Inhaltliche Metadaten

Fachgebiet Informatik
Abstract Using NuPIC to get live traffic anomalies.

Zugehöriges Material

Ähnliche Filme

Loading...
Feedback