Phantom Embeddings: Using Embeddings Space for Model Regularization in Deep Neural Networks.
Mofassir ul Islam ArifMohsan JameelJosif GrabockaLars Schmidt-ThiemePublished in: LWDA (2020)
Keyphrases
- neural network
- computational model
- neural network model
- experimental data
- probabilistic model
- three dimensional
- vector space
- mathematical model
- maximum likelihood
- probability distribution
- cost function
- pattern recognition
- support vector
- objective function
- conceptual model
- prior information
- high level
- neural network is trained