4 ms·
These concepts are orthogonal so that’s probably expected.
by constantly 3y ago
These concepts are orthogonal so that’s probably expected.
- slowhadoken 3y agoorthogonal? Neural networks are a layered graph structure. When they talk about "parameters" they're referring to nodes and edges.
- constantly 3y agoTikTok is backed by complex algorithms to ensure reliability of data transfer and yet I bet you don’t chastise people recording the latest TikTik dance trend for not knowing the fundamentals of TCP/IP. When they talk about sending a video they’re referring to TCP/IP. I bet they can’t even talk about elliptic curve cryptography!
- slowhadoken 3y agoApples and oranges, bud.
- constantly 3y agoSounds like you’re gatekeeping to be honest. People don’t need to know irrelevant concepts or hear stump the chump questions. If that’s how you approached it, great! There are lots of successful approaches and people shouldn’t be chastised because they took a different path or are interested in a different facet of the topic.
- rg111 3y agoHonestly speaking, never had any professor or researcher ever introduce NNs through graph theory. GP here is out of their depth.
- slowhadoken 3y agoLmfao What college did you attend?
- slowhadoken 3y agoYou’re not communicating in good faith. Thrash all you like.
- rg111 3y agoDude, that's not how anyone teaches NNs or learns NNs. You don't need graph theory to understand NNs at all. You need Linear Algebra and Differential Calculus, though. You do need graph theory to understand and do Graph Neural Networks. But that's a subfield of modern AI and many AI researchers don't know/study/research Graph NNs at all. You don't actually need to know graph theory to do RL, Vision, ANNs, NLP, etc. unless explicitly needed in your research/job.
- slowhadoken 3y agoA neural network is a graph structure. It’s not rocket science. It’s basic computer science data structure stuff. If you know linear algebra and differential calculus understanding what a graph is is trivial.