Regularization in Network Optimization via Trimmed Stochastic Gradient Descent With Noisy Label.
Kensuke NakamuraBong-Soo SohnKyoung-Jae WonByung-Woo HongPublished in: IEEE Access (2022)
Keyphrases
- stochastic gradient descent
- stochastic gradient
- matrix factorization
- least squares
- loss function
- step size
- random forests
- regularization parameter
- early stopping
- weight vector
- support vector machine
- importance sampling
- online algorithms
- missing data
- multiple kernel learning
- random forest
- learning algorithm
- bayesian framework
- linear combination
- collaborative filtering
- training data