SVMDense Example 4: Sigmoid Model - Teradata Vantage

Machine Learning Engine Analytic Function Reference

Product
Teradata Vantage
Release Number
8.00
1.0
Published
May 2019
Language
English (United States)
Last Update
2019-11-22
dita:mapPath
blj1506016597986.ditamap
dita:ditavalPath
blj1506016597986.ditaval
dita:id
B700-4003
lifecycle
previous
Product Category
Teradata Vantage™

SQL Call

DROP TABLE densesvm_iris_sigmoid_model;

SELECT * FROM SVMDense (
  ON svm_iris_train AS InputTable
  OUT TABLE ModelTable (densesvm_iris_sigmoid_model)
  USING
  IDColumn ('id')
  InputColumns ('[1:4]')
  ResponseColumn ('species')
  Cost (1)
  Bias (0)
  KernelFunction ('sigmoid')
  Gamma (0.1)
  HashBits (512)
  SubspaceDimension (120)
  MaxStep (30)
  Seed (1)
) AS dt;

Output

message
Model table is created successfully
The model is trained with 120 samples and 512 unique attributes with hash projection
There are 3 different classes in the training set
The model is converged after 18 steps with epsilon 0.01, the average value of the loss function for the training set is 55.832581089711596
The corresponding training parameters are cost:1.0 bias:0.0

Only models with RBF and Sigmoid kernels have converged. This may mean that the true boundaries in the data set are hard to capture with a linear or polynomial model.