Support Vector Machines | Machine Learning

Live Engine

Select Topic

easySupport Vector Machines

An SVM with a linear kernel is trained on a binary classification problem. After training, you discover that removing 90% of training samples — those far from the decision boundary — does not change the model's predictions at all. What does this reveal about how SVMs define their decision boundary?

Live Engine

Select Topic

easySupport Vector Machines

An SVM with a linear kernel is trained on a binary classification problem. After training, you discover that removing 90% of training samples — those far from the decision boundary — does not change the model's predictions at all. What does this reveal about how SVMs define their decision boundary?