I spent most of today trying to come up with a convincing argument for caring about SVMs.
What I thought: Until the late 2000s, SVMs were the dominant model for tasks like image recognition.
What is actually true: Deep learning has outperformed SVMs basically the entire time.
Cozy little instance for anyone who is comfy being part of something called "prettyqueer.online". We have a small cap right now and don't plan on increasing it, so if you know one of the admins, you might ask for an invite but otherwise, it might be best to find another home. <3