Debiasing Averaged Stochastic Gradient Descent to handle missing values.
Aude SportisseClaire BoyerAymeric DieuleveutJulie JossePublished in: NeurIPS (2020)
Keyphrases
- missing values
- stochastic gradient descent
- missing data
- matrix factorization
- loss function
- least squares
- step size
- incomplete data
- support vector machine
- random forests
- regularization parameter
- feature values
- high dimensional data
- weight vector
- hyperparameters
- image sequences
- multiple kernel learning
- importance sampling
- online algorithms
- low rank
- feature vectors