Using SamplingStrategy with smote, adasyn or borderline Values - Teradata Vantage

Used known iris dataset, where species values are changed to 1, 2, and 3.

Use the following code to create the table and insert the rows used in this example:

CREATE MULTISET TABLE iris_enc_sample ,FALLBACK ,
     NO BEFORE JOURNAL,
     NO AFTER JOURNAL,
     CHECKSUM = DEFAULT,
     DEFAULT MERGEBLOCKRATIO,
     MAP = TD_MAP1
     (
      id INTEGER,
      sepal_length FLOAT,
      sepal_width FLOAT,
      petal_length FLOAT,
      petal_width FLOAT,
      species INTEGER)
PRIMARY INDEX ( id );
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (78,6.0,2.9,4.5,1.5,2);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (141,6.9,3.1,5.1,2.3,3);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (97,6.2,2.9,4.3,1.3,2);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (13,4.3,3.0,1.1,0.1,1);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (112,6.8,3.0,5.5,2.1,3);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (89,5.5,2.5,4.0,1.3,2);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (7,5.0,3.4,1.5,0.2,1);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (85,6.0,3.4,4.5,1.6,2);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (5,5.4,3.9,1.7,0.4,1);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (83,6.0,2.7,5.1,1.6,2);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (39,5.1,3.4,1.5,0.2,1);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (35,5.0,3.2,1.2,0.2,1);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (94,5.6,2.7,4.2,1.3,2);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (27,5.2,3.5,1.5,0.2,1);
INSERT INTO iris_enc_sample (id,sepal_length,sepal_width,petal_length,petal_width,species) VALUES (111,6.4,2.7,5.3,1.9,3);

TD_SMOTE SQL Call Using SamplingStrategy with smote

SELECT * FROM TD_SMOTE (
      ON iris_enc_sample AS InputTable PARTITION BY ANY
      USING
      NumberOfNeighbors(5)
      IDColumn('id')
      MinorityClass('3')
      ResponseColumn('species')
      InputColumns('sepal_length', 'sepal_width', 'petal_length', 'petal_width')
      OversamplingFactor(2)
      SamplingStrategy('smote')
) AS dt;

TD_SMOTE SQL Call Using SamplingStrategy with adasyn

SELECT * FROM TD_SMOTE (
      ON iris_enc_sample AS InputTable PARTITION BY ANY
      USING
      NumberOfNeighbors(5)
      IDColumn('id')
      MinorityClass('3')
      ResponseColumn('species')
      InputColumns('sepal_length', 'sepal_width', 'petal_length', 'petal_width')
      OversamplingFactor(2)
      SamplingStrategy('adasyn')
) AS dt;

TD_SMOTE SQL Call Using SamplingStrategy with borderline

SELECT * FROM TD_SMOTE (
      ON iris_enc_sample AS InputTable PARTITION BY ANY
      USING
      NumberOfNeighbors(10)
      IDColumn('id')
      MinorityClass('3')
      ResponseColumn('species')
      InputColumns('sepal_length', 'sepal_width', 'petal_length', 'petal_width')
      OversamplingFactor(2)
      SamplingStrategy('borderline')
) AS dt;