When should you avoid using SVM?

Avoid SVM for extremely large datasets or highly noisy overlapping classes.

SVM becomes inefficient with large datasets and may struggle when classes overlap heavily. In such cases, models like neural networks or tree-based methods may perform better.