Accent Classification is Challenging but Pre-training Helps: a case study with novel Brazilian Portuguese datasets.
Ariadne MatosGustavo AraújoArnaldo Cândido JúniorMoacir PontiPublished in: PROPOR (2024)
Keyphrases
- training dataset
- brazilian portuguese
- benchmark datasets
- training set
- training phase
- classification accuracy
- training samples
- supervised learning
- uci machine learning repository
- classification algorithm
- linear svm
- uci repository
- pattern classification
- support vector
- support vector machine
- automatic classification
- classification method
- pattern recognition
- classification performances
- avoid overfitting
- associative classifiers
- training process
- multi layer perceptron
- machine learning algorithms
- class labels
- speech recognition
- machine learning
- artificial intelligence
- feature extraction
- labelled data
- sufficient training data
- restricted boltzmann machine
- neural network
- feature selection
- decision trees
- classification scheme
- test bed
- benchmark data sets
- classification rules
- data sets
- training examples
- feature vectors
- image classification
- feature set