A Penalized Method for the Predictive Limit of Learning.
Jie DingEnmao DiaoJiawei ZhouVahid TarokhPublished in: ICASSP (2018)
Keyphrases
- learning algorithm
- learning scheme
- high precision
- cost function
- learning process
- prior knowledge
- unsupervised learning
- segmentation method
- synthetic data
- clustering method
- experimental evaluation
- supervised learning
- inductive inference
- high accuracy
- computational cost
- pairwise
- similarity measure
- probabilistic model
- dynamic programming
- online learning
- em algorithm
- model selection
- significant improvement
- theoretical analysis
- active learning
- detection method
- learning systems
- preprocessing
- machine learning
- data sets