A Combined Task Analysis Method for Data Selection in Mandarin Isolated Word Recognition System.
Z. Y. HeZ. G. WangW. LiJ. WuPublished in: ISCSLP (2008)
Keyphrases
- synthetic data
- noisy data
- data sets
- preprocessing
- test data
- missing values
- input data
- missing data
- data analysis
- prior knowledge
- training samples
- clustering method
- classification accuracy
- data sources
- high quality
- feature extraction
- statistical methods
- prior information
- segmentation method
- spectral clustering
- recognition algorithm
- correlation analysis
- data processing
- training data
- data structure
- significant improvement
- cost function
- dynamic programming
- support vector machine
- data collection
- high accuracy
- speech recognition
- feature set
- computational complexity
- data mining techniques
- data quality
- selection strategy
- user input
- data mining