|
|
|
|
|
|
by hzay
716 days ago
|
|
|
Visualizing 3D matrix multiplications, and getting comfortable with it. Then there's basic calculus in understanding gradient descent. Can't think of any other advanced math that was necessary to grok the innermost workings of most models today. Source: I won a silver medal in a kaggle competition after 6 months of ML self-learning. |
|
|