Boosting the generalization ability of Vis-NIR-spectroscopy-based regression models through dimension reduction and transfer learning.
Xiaoli LiZexi LiXufeng YangYong HePublished in: Comput. Electron. Agric. (2021)
Keyphrases
- generalization ability
- transfer learning
- dimension reduction
- regression model
- vis nir
- learning algorithm
- partial least squares
- ensemble methods
- feature selection
- learning tasks
- machine learning algorithms
- model selection
- machine learning
- principal component analysis
- labeled data
- feature extraction
- reinforcement learning
- high dimensional
- active learning
- low dimensional
- support vector
- linear discriminant analysis
- text classification
- multi task
- support vector machine
- high dimensional data
- text categorization
- prediction accuracy
- prediction model
- feature space
- ls svm
- variable selection
- singular value decomposition
- cluster analysis
- high dimensionality
- support vector machine svm
- benchmark datasets
- decision trees
- bp neural network
- training data
- supervised learning
- semi supervised learning
- collaborative filtering
- unsupervised learning
- data analysis
- feature set
- feature vectors
- classification accuracy
- machine learning methods
- semi supervised
- fault diagnosis
- back propagation
- text mining