The Directional Bias Helps Stochastic Gradient Descent to Generalize in Kernel Regression Models.
Yiling LuoXiaoming HuoYajun MeiPublished in: ISIT (2022)
Keyphrases
- regression model
- stochastic gradient descent
- multiple kernel learning
- least squares
- matrix factorization
- loss function
- step size
- random forests
- linear svm
- linear model
- kernel methods
- kernel function
- support vector
- gaussian process
- weight vector
- model selection
- support vector machine
- logistic regression
- online algorithms
- convergence rate
- linear combination
- regularization parameter
- feature space
- text categorization
- learning rate
- machine learning algorithms
- learning problems
- collaborative filtering
- probabilistic model