Classification regularized dimensionality reduction improves ultrasound thyroid nodule diagnostic accuracy and inter-observer consistency.
Wenli DaiYan CuiPeiyi WangHao WuLei ZhangYeping BianYingying LiYutao LiHairong HuJiaqi ZhaoDong XuDexing KongYajuan WangLei XuPublished in: Comput. Biol. Medicine (2023)
Keyphrases
- classification accuracy
- dimensionality reduction
- feature extraction
- accuracy rate
- high dimensionality
- dimensionality reduction methods
- pattern recognition
- ultrasound images
- lung nodules
- pattern classification
- classification rate
- machine learning
- support vector
- fold cross validation
- high accuracy
- support vector machine svm
- roc curve
- multiple criteria linear programming
- training set
- feature space
- support vector machine
- feature vectors
- data representation
- training data
- model selection
- microarray
- principal component analysis
- medical diagnosis
- image processing
- feature selection
- neural network
- data sets