holdout set
[/ˈhoʊldaʊt sɛt/]
nounpl: holdout sets
conjunto de retenção
1. In machine learning and statistics, a subset of data that is set aside and not used during model training, reserved for final evaluation and validation of model performance
The researcher divided the dataset into a training set and a holdout set to evaluate the model's generalization ability.
O pesquisador dividiu o conjunto de dados em um conjunto de treinamento e um conjunto de retenção para avaliar a capacidade de generalização do modelo.
2. A portion of available data deliberately withheld from the training process to provide an unbiased estimate of model performance on unseen data
Using a holdout set helps prevent overfitting and ensures the model performs well on new, unseen examples.
Usar um conjunto de retenção ajuda a prevenir o sobreajuste e garante que o modelo tenha um bom desempenho em exemplos novos e não vistos.
This is a technical term primarily used in data science, machine learning, and statistics communities. It emerged from English-speaking academic and tech sectors and is widely adopted internationally. In Brazil, it's commonly used in tech hubs like São Paulo and among data science professionals, though Portuguese translations vary. The concept reflects modern practices in machine learning where data separation is crucial for building reliable predictive models.
Look up more words on Fala2Me
The free English-Portuguese dictionary with real Brazilian accents, NYC slang, conjugator and more
Open Fala2Me →