Dimensionality Reduction for Using High-Order n-Grams in SVM-Based Phonotactic Language Recognition.
Mikel PeñagarikanoAmparo VaronaLuis Javier RodríguezGermán BordelPublished in: INTERSPEECH (2011)
Keyphrases
- high order
- n gram
- dimensionality reduction
- feature extraction
- language specific
- character n grams
- higher order
- language model
- low order
- language independent
- variable length
- pairwise
- text classification
- pattern recognition
- bag of words
- part of speech
- lower order
- support vector
- support vector machine
- action recognition
- principal component analysis
- data representation
- low dimensional
- high dimensional
- inside outside algorithm
- markov random field
- subspace learning
- word level
- image classification
- transfer learning
- web documents
- computer vision
- relevance feedback
- knn
- data points
- probabilistic model
- computational complexity
- natural language
- face recognition