Cross Entropy in Deep Learning of Classifiers Is Unnecessary - ISBE Error is All You Need.
Wladyslaw SkarbekPublished in: CoRR (2023)
Keyphrases
- deep learning
- cross entropy
- error function
- log likelihood
- unsupervised learning
- maximum likelihood
- machine learning
- training data
- decision trees
- support vector
- language modeling
- mental models
- learning rate
- feature set
- neural network
- weakly supervised
- data sets
- training set
- ranking functions
- activation function
- supervised learning
- pairwise
- feature selection
- natural language