meta-classifier
[/ˈmɛtə ˈklæsɪfaɪər/]
nounmasculinepl: meta-classifiers
meta-classificador
1. A machine learning model that learns to combine or select predictions from multiple base classifiers to improve overall classification performance
The ensemble used a meta-classifier to optimally weight the outputs from five different neural networks.
O ensemble usou um meta-classificador para ponderar otimamente as saídas de cinco redes neurais diferentes.
2. In stacked generalization, a classifier trained on the outputs of other classifiers rather than on raw features
A meta-classifier in a stacking framework learns patterns in how base models make errors.
Um meta-classificador em um framework de stacking aprende padrões em como os modelos base cometem erros.
3. A second-level classifier that determines the most appropriate base classifier for a given input
The meta-classifier algorithm decided which specialized model to use for each new data instance.
O algoritmo meta-classificador decidiu qual modelo especializado usar para cada nova instância de dados.
This is a specialized technical term used primarily in machine learning and artificial intelligence research communities. It is equally recognized in both Brazilian Portuguese and European Portuguese technical literature. The term is relatively recent in origin (early 2000s) and reflects the internationalization of machine learning terminology across Portuguese-speaking regions. Usage is identical in academic and industrial settings worldwide.
Look up more words on Fala2Me
The free English-Portuguese dictionary with real Brazilian accents, NYC slang, conjugator and more
Open Fala2Me →