Login / Signup
Gradient Descent with Early Stopping is Provably Robust to Label Noise for Overparameterized Neural Networks.
Mingchen Li
Mahdi Soltanolkotabi
Samet Oymak
Published in:
CoRR (2019)
Keyphrases
</>
label noise
early stopping
neural network
active learning
training data
data mining
noise model
noise tolerant
data sets
image processing
cost function
worst case
machine learning algorithms
stochastic gradient descent