Improvements to color image and machine learning based thin-film nitrate sensor performance prediction: New texture features, repeated cross-validation, and auto-tuning of hyperparameters.
Xihui WangYe MiAli ShakouriGeorge T.-C. ChiuJan P. AllebachPublished in: Color Imaging: Displaying, Processing, Hardcopy, and Applications (2022)
Keyphrases
- cross validation
- hyperparameters
- texture features
- thin film
- model selection
- machine learning
- color images
- support vector
- regularization parameter
- parameter optimization
- gaussian process
- gray level
- image features
- distance measure
- grid search
- feature vectors
- color space
- training set
- em algorithm
- feature extraction
- support vector machine
- learning algorithm
- sample size
- neural network
- data mining
- feature selection
- feature set
- active learning
- prior information
- color information
- random sampling
- learning tasks
- feature space
- higher order
- multiscale
- parameter estimation
- decision trees
- image processing
- generative model
- data sets