Global Climate Models (GCMs) are the primary tool to simulate climate
ev...
The parameter space for any fixed architecture of feedforward ReLU neura...
Sharpness-aware minimization (SAM) was proposed to reduce sharpness of m...
Climate simulations are essential in guiding our understanding of climat...
Climate change is a major driver of biodiversity loss, changing the
geog...
Machine learning algorithms for parsing remote sensing data have a wide ...
Training a neural network requires choosing a suitable learning rate,
in...
Mitigating the climate crisis requires a rapid transition towards lower
...
Policies produced by deep reinforcement learning are typically character...
ImageNet-1k is a dataset often used for benchmarking machine learning (M...
The availability of reliable, high-resolution climate and weather data i...
Current deep learning approaches have shown good in-distribution
general...
Labeled datasets for agriculture are extremely spatially imbalanced. Whe...
Numerical simulations of Earth's weather and climate require substantial...
Many experts argue that the future of artificial intelligence is limited...
Large optimization problems with hard constraints arise in many settings...
Assessing the complexity of functions computed by a neural network helps...
The output of a neural network depends on its parameters in a highly
non...
Climate change is one of the greatest challenges facing humanity, and we...
The success of deep networks has been attributed in part to their
expres...
It is well-known that the expressivity of a neural network depends on it...
Pixel-accurate tracking of objects is a key element in many computer vis...
Continual learning is the problem of learning new tasks or knowledge whi...
Neural network optimization is often conceptualized as optimizing parame...
We investigate the effects of initialization and architecture on the sta...
Deep learning algorithms for connectomics rely upon localized classifica...
Deep neural networks trained on large supervised datasets have led to
im...
It is well-known that neural networks are universal approximators, but t...
The field of connectomics faces unprecedented "big data" challenges. To
...
We show how the success of deep learning could depend not only on mathem...
We introduce a new problem in the approximate computation of Gröbner bas...