On Improving the Algorithm-, Model-, and Data- Efficiency of Self-Supervised Learning.
Yun-Hao CaoJianxin WuPublished in: CoRR (2024)
Keyphrases
- input data
- learning algorithm
- learned models
- prior knowledge
- learning phase
- learning models
- probabilistic model
- experimental data
- automatically learned
- data sets
- theoretical analysis
- cost function
- objective function
- prior information
- mathematical model
- training data
- recognition algorithm
- parameter estimation
- bayesian methods
- detection algorithm
- data reduction
- data samples
- learning process
- noisy data
- synthetic datasets
- background knowledge
- dynamic programming
- single scan
- classification algorithm
- tree structure
- estimation algorithm
- learning scheme
- accurate models
- computational complexity
- data analysis
- data structure
- preprocessing
- active learning
- hidden variables
- segmentation algorithm
- data points
- high efficiency
- neural network
- network structure
- kalman filter
- energy function
- labeled data
- em algorithm
- expectation maximization
- probability distribution
- data sources
- similarity measure
- bayesian framework
- missing data
- high dimensional data
- clustering method
- model free
- learning mechanism
- supervised learning