A New Data Augmentation Method for Intent Classification Enhancement and its Application on Spoken Conversation Datasets.
Zvi KonsAharon SattHong-Kwang KuoSamuel ThomasBoaz CarmeliRon HooryBrian KingsburyPublished in: ICASSP (2022)
Keyphrases
- synthetic data
- classification method
- test data
- classification accuracy
- data sets
- training samples
- preprocessing
- support vector machine svm
- input data
- ordinal data
- input vectors
- machine learning methods
- binary and multi class
- noisy data
- statistical methods
- prior knowledge
- database
- training data
- similarity measure
- large scale data sets
- gene expression microarray data
- dimensionality reduction methods
- sampling methods
- support vector machine
- pairwise
- training set
- clustering method
- training dataset
- data points
- feature extraction
- original data
- raw data
- missing values
- probabilistic model
- pattern classification
- classification process
- dimensional data
- multi class
- decision rules
- cross validation
- missing data
- feature set
- data structure
- high dimensional data
- data sources
- receiver operating characteristic curves
- machine learning
- image processing
- high dimensional datasets
- data mining techniques
- image classification
- classification trees
- text classification
- bayesian methods
- data analysis
- extracted features
- benchmark data sets
- feature subset
- classification algorithm
- pattern recognition