Skip to content

johnvorsten/point_categorizer

Repository files navigation

Packages:

  1. clustering
    Clustering algorithms and modules to cluster points based on similarity. After the points are clustered, they are loaded into mongodb to be used in the MIL categorization section of this project. TODO : Move some of the database pipeline modules OUT of clustering, pipelines & mongo

  1. data Includes .csv files for clustering, manual cleaning of some data, tensorflow TF record protos, vocabulary files, and a master .csv file of extracted databases from .mdf to .csv

  1. MILCategorization Multiple instance categorization for bags of clustered data. Optionally, use this on databases segmented based on controller instead of clustered databases.

  1. ranking Holds models for tensorflow ranking models. Includes rankign of optimal hyperparameters for clustering databases based on points similarity

About

Automatically label BAS

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published