home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: misc.invest
- Path: sparky!uunet!haven.umd.edu!darwin.sura.net!spool.mu.edu!uwm.edu!linac!att!cbnews!lib
- From: lib@cbnews.cb.att.com (Lib)
- Subject: Re: Predictions
- Organization: AT&T
- Date: Tue, 26 Jan 1993 13:25:21 GMT
- Message-ID: <1993Jan26.132521.18105@cbnews.cb.att.com>
- Keywords: Predictions
- Lines: 18
-
- In article <1k2909INNmer@lynx.unm.edu> scc26mf@phobos.unm.edu (Mark Fleharty) writes:
- >
- >For those of you who want to know how my program works I will explain it in
- >moderate detail now.
- >
- >My program works by example. It compares todays data with data from the past.
- >If it finds data in the past that matches well based on volitility. It then
- >matches todays data with the data that preceedes the example data. Thus it
- >comes up with a signal.
-
- Let's see if I understand.
-
- You look at recent trading data, find a time period
- where the data looks similar, look at the data which PRECEEDS that
- period and use that preceeding data to predict what will come AFTER
- the current period. Not particularly intuitive but if it works
- then great.
-
-