2 papers
cs.LG2020
Bayesian Sampling Bias Correction: Training with the Right Loss Function
L. Le Folgoc, V. Baltatzis, A. Alansary +8
We derive a family of loss functions to train models in the presence of sampling bias. Examples are when the prevalence of a pathology differs from its sampling rate in the trainin…
cs.LG2019
Exploiting Uncertainty of Loss Landscape for Stochastic Optimization
Vineeth S. Bhaskara, Sneha Desai
We introduce novel variants of momentum by incorporating the variance of the stochastic loss function. The variance characterizes the confidence or uncertainty of the local feature…