Despite strong empirical performance for image classification, deep neur...
This paper studies the curious phenomenon for machine learning models wi...
While cross entropy (CE) is the most commonly used loss to train deep ne...
When training deep neural networks for classification tasks, an intrigui...
Recently, over-parameterized deep networks, with increasingly more netwo...
State-of-the-art subspace clustering methods are based on self-expressiv...
This work attempts to provide a plausible theoretical framework that aim...
We provide the first global optimization landscape analysis of
Neural Co...
Normalization techniques have become a basic component in modern
convolu...
Current deep learning architectures suffer from catastrophic forgetting,...
This work attempts to interpret modern deep (convolutional) networks fro...
Subspace clustering is an unsupervised clustering technique designed to
...
Initialization, normalization, and skip connections are believed to be t...
Recent advances have shown that implicit bias of gradient descent on
ove...
To learn intrinsic low-dimensional structures from high-dimensional data...
In over two decades of research, the field of dictionary learning has
ga...
Finding a small set of representatives from an unlabeled dataset is a co...
Subspace clustering methods based on expressing each data point as a lin...
State-of-the-art subspace clustering methods are based on self-expressiv...
The classical bias-variance trade-off predicts that bias decreases and
v...
Given an overcomplete dictionary A and a signal b = Ac^* for some sparse...
Subspace clustering methods based on data self-expression have become ve...
Semiparametric regression offers a flexible framework for modeling non-l...
Sparse subspace clustering (SSC) is a state-of-the-art method for segmen...
Many computer vision tasks involve processing large amounts of data
cont...
Subspace clustering refers to the problem of segmenting data drawn from ...
State-of-the-art subspace clustering methods are based on expressing eac...
Subspace clustering methods based on ℓ_1, ℓ_2 or nuclear norm
regulariza...