Statistical Learning
Acceleration with a Ball Optimization Oracle Y air Carmon
In the introduction we discuss exact oracles for simplicity, but our results account for inexactness. Our results hold for any weighted Euclidean (semi)norm, i.e.,
Acceleration with a Ball Optimization Oracle Y air Carmon
In the introduction we discuss exact oracles for simplicity, but our results account for inexactness. Our results hold for any weighted Euclidean (semi)norm, i.e.,
Non-convex online learning via algorithmic equivalence
We study an algorithmic equivalence technique between non-convex gradient descent and convex mirror descent.
Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods
We collect and introduce diverse non-homophilous datasets from a variety of application areas that have up to 384x more nodes and 1398x more edges than prior datasets.
Supplementary Material for The Origins and Prevalence of Texture Bias in Neural Networks
Higher learning rates produce greater shape bias. As shown in Figure A.1, higher values of learning rate and weight decay were associated with greater We found that random-crop augmentation biases models towards texture (Section 5). We did not change the aspect ratio or other data augmentation settings. Networks with limited receptive fields learn texture more easily than shape. Shape is persistently more decodable through the convolutional layers of AlexNet than is texture, which rises through them.
