Skip to main content

train_binary_classifier

Tutorial and use case​

Syntax​

train_binary_classifier(<table_name>, 
                        <target_column_name>,
                        <excluded_column_name_list>)

Arguments​

<table_name> (string) - the data table's name.

<target_column_name> (string) - the name of the target column in the data table.

<excluded_column_name_list> (list) - a list of the column names that should be excluded (ignored) from training.

Algorithm details​

12 ML algorithms are used in the training process. They are:

  • Logistic Regression
  • Naive Bayes
  • Linear Discriminant Analysis
  • Gradient Boosting Classifier
  • Ada Boost Classifier
  • Quadratic Discriminant Analysis
  • Ridge Classifier
  • Light Gradient Boosting Machine
  • Random Forest Classifier
  • K Neighbors Classifier
  • Extra Trees Classifier
  • Decision Tree Classifier
  • Support Vector Machine

Example​

SQL statement​

CALL TRAIN_BINARY_CLASSIFIER('CUSTOMER', 'CHURN', ['CUSTOMERID']);

Description​

Train a binary classifier using data in "CUSTOMER" table, where column "CHURN" is the target, and "CUSTOMERID" column should be ignored as it's not a good feature.

Result​

All historical and forecasted data is materialized as a new table.

A model performance dashboard is auto-generated for checking modeling accuracy, discovered trends, and predictions.

Prerequisite​

Input data format​

Input data must include at least one feature and one target column. And the target column should only contain 2 values.