1.0 - 8.00 - GMM Input - Teradata Vantage

Teradata® Vantage Machine Learning Engine Analytic Function Reference

Teradata Vantage
Release Number
Release Date
May 2019
Content Type
Programming Reference
Publication ID
English (United States)
Table Description
InputTable Contains input data to cluster.
init_params [Optional] Table or view that contains initial values for weight, mean, and covariance of each cluster. If this table or view is empty, function terminates without executing.
If you do not specify a table or view, you must use query ON (SELECT 1), which causes function to determine initial values as follows:
Characteristic Initial Value
weight Same for all clusters.
mean Selected from multivariate standard normal distribution centered at origin.
covariance Covariance matrix computed by function and assigned to each cluster.

InputTable Schema

Column Data Type Description
id Any Leftmost column. Data point identifier.
dim_n Any numeric SQL data type [Column appears once for each dimension.] Data in dimension i.

init_params Table Schema

Column Data Type Description
weight Any numeric SQL data type Initial weight of cluster.
dim_n Any numeric SQL data type [Column appears once for each dimension.] Initial mean of cluster.
covariance VARCHAR Initial covariance of cluster. Possible values depend on CovarianceType.
CovarianceType Possible Values
'spherical' Positive numeric value (for example, 1.0.)
'diagonal' JSON representation of DOUBLE PRECISION array (for example, [1.0,2.0,3.0,4.0])
'tied' or 'full' JSON representation of two-dimensional DOUBLE PRECISION array (for example, [[1.0,2.0],[2.0,4.0]])