home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!darwin.sura.net!wupost!waikato.ac.nz!canterbury.ac.nz!math!jsv
- Newsgroups: sci.math.stat
- Subject: Re: Help - A TIME SERIES question
- Message-ID: <BxzoMy.7Iz@cantua.canterbury.ac.nz>
- From: jsv@math.canterbury.ac.nz (Julian Visch)
- Date: Fri, 20 Nov 1992 00:42:34 GMT
- References: <37667@uflorida.cis.ufl.edu>
- Organization: Department of Mathematics, University of Canterbury
- Nntp-Posting-Host: sss330.canterbury.ac.nz
- Lines: 43
-
- In article <37667@uflorida.cis.ufl.edu>, brv@reef.cis.ufl.edu (Balaji Venkatraman) writes:
- |> Hi,
- |>
- |> I am a graduate student in computer science with not much back ground
- |> in statistics, specifically TIME SERIES analysis.
- |>
- |> As part of my work, i am trying to analyze the behavior of computer
- |> network traffic on the campus backbone.
- |> Each record in the data set consists of (among others):
- |> 1. Packet arrival time
- |> 2. Packet size
- |> 3. Packet type.
- |>
- |> I would like to fit a model to the data set so that i can predict:
- |> 1. Arrival of packets on the network
- |> 2. Packet size
- |>
- |> Another important point (in my view) is to be able to characterize the
- |> "Bursty" nature of traffic.
- |>
- |> I have done considerable measurements and analysis on the traffic from
- |> a networks viewpoint (and published it) and i would like to back it up
- |> with a "good" time series model.
- |>
- |> I would appreciate any help and guidance on this problem. If anyone else
- |> is working on a similar problem, maybe we could continue this discussion
- |> (offline).
- |> Please post your replies or send me E-mail.
- |>
- |> Thanks in advance.
- |>
- |> Balaji R. Venkatraman Internet: brv@cis.ufl.edu
- |> Computer and Information Sciences Phone : (904) 392-1491, 392-1200 (office)
- |> University of Florida (904) 372-3135 (home)
- |> Gainesville, FL 32611. Fax : (904) 392-1220
- |>
-
- If you trying to fit an A.R.I.M.A model
- There is a book by Christopher Chatfield called
- "The analysis of time series : an introduction"
- which I found quite useful
-
- Otherwise you might try reading books on queueing theory.
-