Wordreg: Mitigating the Gap between Training and Inference with Worst-Case Drop Regularization.
Jun XiaGe WangBozhen HuCheng TanJiangbin ZhengYongjie XuStan Z. LiPublished in: ICASSP (2023)
Keyphrases
- worst case
- structured prediction
- early stopping
- training process
- lower bound
- training algorithm
- stochastic gradient descent
- np hard
- training samples
- bayesian networks
- greedy algorithm
- prior information
- special case
- machine learning
- test set
- error bounds
- training examples
- approximation algorithms
- probabilistic inference
- supervised learning
- upper bound
- training phase
- parameter selection
- average case
- training set
- restricted boltzmann machine