Influence of Varying Training Set Composition and Size on Support Vector Machine-Based Prediction of Active Compounds.
Raquel Rodríguez-PérezMartin VogtJürgen BajorathPublished in: J. Chem. Inf. Model. (2017)
Keyphrases
- training set
- training data
- prediction accuracy
- data sets
- prediction model
- wet lab
- test set
- supervised learning
- active learning
- prediction error
- molecular structure
- training samples
- prediction algorithm
- experimental data
- computational complexity
- nearest neighbor
- training set size
- active appearance models
- svm classifier
- high dimensional
- classification accuracy
- class distribution
- web service composition
- factors influencing
- individual differences
- drug discovery
- training examples
- support vector machine
- support vector