Evaluating the Impact of Loss Function Variation in Deep Learning for Classification.
Simon DrägerJannik DunkelauPublished in: CoRR (2022)
Keyphrases
- loss function
- deep learning
- support vector
- hinge loss
- machine learning
- unsupervised learning
- pattern recognition
- support vector machine
- restricted boltzmann machine
- pairwise
- image classification
- training set
- text classification
- weakly supervised
- decision trees
- unsupervised feature learning
- risk minimization
- feature space
- training samples
- feature extraction
- supervised learning
- multi class
- feature selection
- learning algorithm