Differences between revisions 1 and 9 (spanning 8 versions)
Revision 1 as of 2005-03-23 23:40:21
Size: 753
Editor: yakko
Comment: .
Revision 9 as of 2005-08-03 14:23:51
Size: 1042
Editor: yakko
Comment:
Deletions are marked like this. Additions are marked like this.
Line 1: Line 1:
= Dynamic Max Count = = Dynamic Max Coun =
Line 3: Line 3:
This contains the ideas and notes for a Dynamic Max Count or Max-in-time agregate operator This contains the ideas and notes for a Dynamic Max Count (Dynamic
Max-in-time) aggregate operator
Line 7: Line 8:
Instead of using Hyper-buckets that have descrete boundaries and densities which can not be updated resonably using the MaxCount ideas, we propose a probabalistic method where by we put probability densities in space. Each probability density will need the following properties: Instead of using Hyper-buckets that have discrete boundaries and
densities which can not be updated reasonably using the
MaxCountProgramNotes ideas, we propose a probabilistic method where
by we put probability densities in space. Each probability density
will need the following properties:
Line 13: Line 18:
      1. A measure of symetry or skew
   1. A multi-dimensional probability function preferably a function that specific types functions as parameters e.g. p(x_u(t),x_l(t),y_u(t),y_l(t)[,z_u(t),z_l(t)])
      1. A measure of symmetry or skew
   1. A multi-dimensional probability function preferably a function that uses types functions as parameters e.g. [[latex2(p(xu(t),xl(t),yu(t),yl(t)[,zu(t),zl(t)]))]]
   1. A theory to ''update'', ''delete'' or ''insert'' points and the distributions based on changes to points.

Based on this last item, we must maintain a database of
4-dimensional points that we index using 4-dimensional, probability
buckets.

Dynamic Max Coun

This contains the ideas and notes for a Dynamic Max Count (Dynamic Max-in-time) aggregate operator

Concept

Instead of using Hyper-buckets that have discrete boundaries and densities which can not be updated reasonably using the MaxCountProgramNotes ideas, we propose a probabilistic method where by we put probability densities in space. Each probability density will need the following properties:

  1. Parameters that define the distribution e.g.
    1. Center location
    2. Spatial size
    3. Standard deviation
    4. A measure of symmetry or skew
  2. A multi-dimensional probability function preferably a function that uses types functions as parameters e.g. latex2(p(xu(t),xl(t),yu(t),yl(t)[,zu(t),zl(t)]))

  3. A theory to update, delete or insert points and the distributions based on changes to points.

Based on this last item, we must maintain a database of 4-dimensional points that we index using 4-dimensional, probability buckets.

DynamicMaxCount (last edited 2020-01-23 22:27:01 by 68)