From the sparse input table svm_iris_input, create a training table (with 80% of the rows) and a testing table (with 20% of the rows):
CREATE TABLE svm_iris_input_train AS SELECT * FROM svm_iris_input WHERE id%5!=0; CREATE TABLE svm_iris_input_test AS SELECT * FROM svm_iris_input WHERE id%5=0;
The testing table is input to the SparseSVMPredictor function. The testing table, which is input to the SparseSVMPredictor function, is Input.
id | species | attribute | value |
---|---|---|---|
1 | setosa | sepal_length | 5.1 |
1 | setosa | sepal_width | 3.5 |
1 | setosa | petal_length | 1.4 |
1 | setosa | petal_width | 0.2 |
2 | setosa | sepal_length | 4.9 |
2 | setosa | sepal_width | 3.0 |
2 | setosa | petal_length | 1.4 |
2 | setosa | petal_width | 0.2 |
3 | setosa | sepal_length | 4.7 |
3 | setosa | sepal_width | 3.2 |
... | ... | ... | ... |