Class Weights vs Resampling

Medium

machine-learning PayPal Robinhood Coinbase

You are training a logistic regression model on an imbalanced dataset (10% positive, 90% negative). A teammate suggests using SMOTE, but you prefer adjusting class weights in the loss function. What is the main advantage of using class weights over resampling?

Multiple Choice

Class weights always produce higher accuracy than resampling methods Class weights modify the loss function without changing the training data, avoiding synthetic sample artifacts Class weights can only be used with logistic regression, not other algorithms Class weights eliminate the need for any hyperparameter tuning

Correct!

Incorrect — try again next time!

Upgrade to Pro to unlock solutions and explanations

Class Weights vs Resampling

Multiple Choice

Related problems

Deep-dive guides