SnowPro Advanced Data Scientist

By Snowflake · Question Mode

QuestionQ40

VerifiedTrain and use machine learning models

This Snowsight chart for a New York City ride-share bike service displays the number of trips made to each destination borough:

Question Image

A Data Scientist wants to create a classifier that predicts the borough most likely to be the destination when a trip begins.

Which techniques should be used to address the class imbalance shown in the Snowsight chart?

Choose two
Explanation

Undersampling the overrepresented Manhattan class reduces its dominance in the training set. SMOTE oversamples the underrepresented non-Manhattan classes by generating synthetic minority-class examples, helping the classifier learn their decision boundaries rather than favoring Manhattan. These approaches are standard methods for balancing multiclass training data.

Learn more

Community Discussion

No comments yet. Be the first to start the discussion!