A Theoretical Analysis on Feature Learning in Neural Networks: Emergence from Inputs and Advantage over Fixed Features.
Zhenmei ShiJunyi WeiYingyu LiangPublished in: CoRR (2022)
Keyphrases
- neural network
- theoretical analysis
- feature set
- feature vectors
- image features
- neural nets
- learning systems
- feature values
- learning process
- feature extraction
- active learning
- single feature
- knowledge acquisition
- significant features
- genetic algorithm
- learning tasks
- additional features
- feedforward neural networks
- multiple features
- learning models
- connectionist systems
- numerical simulations
- supervised learning
- prior knowledge
- learning algorithm
- incremental learning
- feature descriptors
- back propagation
- discriminative features
- low level
- feature analysis
- training set
- object features
- predictive features
- machine learning