Correlation between event and its effect on a time series

For the discussion of math. Duh.

Moderators: gmalivuk, Moderators General, Prelates

csiko
Posts: 2
Joined: Sun Nov 04, 2012 7:51 pm UTC

Correlation between event and its effect on a time series

Postby csiko » Wed Nov 07, 2012 8:43 pm UTC

I have a time series, e.g. the daily number of visitors on my blog. I have a set of events of some class, like the days when I made a new posting. I want to measure the effect of a new posting on the visitors. However, the effect of a new posting is not limited to the same day the posting was made. It could be days later.

So to generalize it: There is a (either discrete time or continous time) series of data points. There is a set of events given by their time instants. All events are believed to have a similar effect on the series, beginning from the event time and lasting for some time t_L.

Is there a statictical method to approach this problem? Or do I need some extra assumptions?

Thanks in advance for any tips, ideas, help on this.

Treatid
Posts: 191
Joined: Thu Sep 13, 2012 11:33 am UTC
Location: United Kingdom
Contact:

Re: Correlation between event and its effect on a time serie

Postby Treatid » Sun Nov 11, 2012 9:48 pm UTC

As a general resource I would imagine that various analysis relating to stock market data could provide some useful algorithms.

To some degree the problem can be considered as a "best fit curve" problem. In this case you would have the underlying data - which presumably is fairly linear - on top of which you have the response to each event. One might guess that the general response to an event will be an initial spike that decays over time - so some form of decay graph and an iterated approach to best fit of parameters.

If your initial event provokes further events (responses in turn provoke discussion...) then a more complex curve may be appropriate.

As mentioned above - for better or worse - curve fitting is a favourite sport among market watchers and there are many well developed algorithms for matching curves to stochastic data.

mouseposture
Posts: 42
Joined: Tue Sep 15, 2009 2:42 am UTC

Re: Correlation between event and its effect on a time serie

Postby mouseposture » Wed Nov 14, 2012 2:45 am UTC


FancyHat
Posts: 341
Joined: Thu Oct 25, 2012 7:28 pm UTC

Re: Correlation between event and its effect on a time serie

Postby FancyHat » Wed Nov 14, 2012 5:55 am UTC

Would autocorrelation be suitable?
I am male, I am 'him'.

User avatar
ConMan
Shepherd's Pie?
Posts: 1690
Joined: Tue Jan 01, 2008 11:56 am UTC
Location: Beacon Alpha

Re: Correlation between event and its effect on a time serie

Postby ConMan » Wed Nov 14, 2012 10:21 pm UTC

FancyHat wrote:Would autocorrelation be suitable?

Well, autocorrelation is better for understanding how a series without any assumed external input behaves. To deal with the effect of external factors, you need to incorporate some form of intervention analysis. While probably overpowered for your purpose, I think X-12-ARIMA from the US Bureau of Labour Statistics includes some suitable intervention analysis methods.
pollywog wrote:
Wikihow wrote:* Smile a lot! Give a gay girl a knowing "Hey, I'm a lesbian too!" smile.
I want to learn this smile, perfect it, and then go around smiling at lesbians and freaking them out.


Return to “Mathematics”

Who is online

Users browsing this forum: No registered users and 18 guests