Here, we introduce Multi-Modal Multitask MIMIC-III (M3) — a dataset and benchmark for evaluating machine learning algorithms in the healthcare domain.
Learning representations that accurately capture long-range dependencies in sequential inputs --- including text, audio, and genomic data --- is a key problem in deep learning.
Ranked #1 on Audio Super-Resolution on VCTK Multi-Speaker
Learning representations that accurately capture long-range dependencies in sequential inputs -- including text, audio, and genomic data -- is a key problem in deep learning.
Ranked #1 on Audio Super-Resolution on Voice Bank corpus (VCTK)
Estimates of predictive uncertainty are important for accurate model-based planning and reinforcement learning.
Methods for reasoning under uncertainty are a key building block of accurate and reliable machine learning systems.
Constraint-based learning reduces the burden of collecting labels by having users specify general properties of structured outputs, such as constraints imposed by physical laws.
Modern machine learning algorithms are often susceptible to adversarial examples — maliciously crafted inputs that are undetectable by humans but that fool the algorithm into producing undesirable behavior.
We introduce a new audio processing technique that increases the sampling rate of signals such as speech or music using deep convolutional neural networks.
Ranked #2 on Audio Super-Resolution on Voice Bank corpus (VCTK)
In user-facing applications, displaying calibrated confidence measures---probabilities that correspond to true frequency---can be as important as obtaining high accuracy.
Tensor factorization arises in many machine learning applications, such knowledge base modeling and parameter estimation in latent variable models.
Although the design of clinical trials has been one of the principal practical problems motivating research on multi-armed bandits, bandit algorithms have never been evaluated as potential treatment allocation strategies.