tag:blogger.com,1999:blog-7831813422886730737.post6827039250169089539..comments2023-10-08T10:44:28.524+03:00Comments on Event Processing Thinking: On CEP and IEPOpher Etzionhttp://www.blogger.com/profile/10791357917675270335noreply@blogger.comBlogger4125tag:blogger.com,1999:blog-7831813422886730737.post-64583366203452939912007-12-07T12:00:00.000+02:002007-12-07T12:00:00.000+02:00Thanks,that was nicely put.Eran.Thanks,<BR/><BR/>that was nicely put.<BR/><BR/>Eran.Eranhttps://www.blogger.com/profile/15249363784380760219noreply@blogger.comtag:blogger.com,1999:blog-7831813422886730737.post-6158958847078452652007-12-06T23:34:00.000+02:002007-12-06T23:34:00.000+02:00Well, there's one big rule of thumb, coming from t...Well, there's one big rule of thumb, coming from the definitions: If there's uncertainly, then it's stochastic. If it's certain, then it's deterministic. :-)<BR/><BR/>Deterministic problems usually become stochastic when you're not exactly sure what an observation means. Taking a variation of traffic lights:<BR/><BR/>If you have a fine-grained sequence of exact locations for a car, an exact location for an intersection and precise knowledge of when the traffic light in the intersection is red and green... well then you can make a deterministic decision on whether the car was illegally crossing the intersection during a red light.<BR/><BR/>It becomes stochastic when you are not sure of the exact meaning of one or more observations. Maybe the sensor that determines the location of the car could be off by up to 20 feet in any direction. Maybe you only have a few observations of the car's location - none of them show the car exactly in the intersection during the red light but some show that it was suspiciously close. As soon as you are unsure of something, this is a candidate for stochastic modeling.Hanshttps://www.blogger.com/profile/03057096345613832279noreply@blogger.comtag:blogger.com,1999:blog-7831813422886730737.post-83713972600641382862007-12-05T10:29:00.000+02:002007-12-05T10:29:00.000+02:00Hello Eran -- I am not sure if there are such auto...Hello Eran -- I am not sure if there are such automatic procedures to decide, there may be some thumb rules.<BR/><BR/>cheers,<BR/><BR/>OpherOpher Etzionhttps://www.blogger.com/profile/17070103285719046013noreply@blogger.comtag:blogger.com,1999:blog-7831813422886730737.post-69914812196289522132007-12-02T11:07:00.000+02:002007-12-02T11:07:00.000+02:00Interesting point, and one which is relevant to ma...Interesting point, and one which is relevant to many other information systems problem. <BR/><BR/>I wonder if there is some general measure which allows us to automatically decide if a given problem will be best treated by stochastic measures or by deterministic ones.Eranhttps://www.blogger.com/profile/15249363784380760219noreply@blogger.com