You are training a logistic regression model to classify emails as spam or not spam. The model is currently classifying too many emails as spam. What would you do to adjust the model?
A.
Decrease the weight of the feature representing the presence of certain keywords.
In binary classification, the decision threshold determines how much predicted probability is required before an observation is assigned to the positive class. Here, spam represents the positive class, and the model is producing too many spam classifications, indicating excessive positive predictions or false positives. Increasing the classification threshold requires a higher predicted probability before an email is labeled as spam, reducing the number of positive classifications. Oracle Machine Learning documentation defines the probability threshold as the decision point used for binary classification and explains that changing this threshold changes true-positive and false-positive behavior. Oracle Docs Altering individual feature weights, iteration counts, or regularization affects model training rather than directly controlling the classification decision boundary. Therefore, increasing the classification threshold is the most appropriate adjustment.
================
Contribute your Thoughts:
Chosen Answer:
This is a voting comment (?). You can switch to a simple comment. It is better to Upvote an existing comment if you don't have anything to add.
Submit