DBSCAN with huge number of points

Geographic Information Systems Asked by Rourich on December 15, 2020

We have a dataset of approximately 30 million new points per month (latitude, longitude, and a measurement associated with that point).

So far, we have applied DBSCAN to generate a monthly map such that the plotted points are:

  • A "centroid" for each group generated by DBSCAN (calculated by us from the points of each group).
  • A centroid color equal to the mean of the measurement associated with all the points used to calculate that centroid.

The problem is that from now instead of making a different map each month with the new data, we want to make a map that includes the new data.

Therefore, I have one question:

Would it make sense to apply DBSCAN using both the new data from one month plus the previously calculated centroids, as if these centroids were a raw dataset? (would be an incremental approximation)

Add your own answers!

Related Questions

Removing JSESSION value from URL of GeoNetwork

0  Asked on July 10, 2021 by vicentgn


How to undo misclicks while drawing a polygon?

2  Asked on July 10, 2021 by simskopf


GDAL import error, broken both python script and Qgis

0  Asked on July 9, 2021 by justin-larking


Defra LiDAR – Convert DTM to PointCloud?

0  Asked on July 9, 2021 by steve-hibbert


Rotating set of polygons as a block

1  Asked on July 9, 2021 by serge-de-gosson-de-varennes


Longitude/latitude to esriMeter

0  Asked on July 9, 2021 by kron_ninja


Map Projection invented by woman

1  Asked on July 8, 2021 by daniele-turchetti


Ask a Question

Get help from others!

© 2022 All rights reserved. Sites we Love: PCI Database, MenuIva, UKBizDB, Menu Kuliner, Sharing RPP, SolveDir