feature scaling
[ˈfiːtʃər ˈskeɪlɪŋ]
noun
normalização de características
1. A preprocessing technique in machine learning that standardizes the range of independent variables or features in a dataset to improve model performance and convergence speed
Feature scaling is essential before applying distance-based algorithms like k-means clustering to ensure that all variables contribute equally to the analysis.
A normalização de características é essencial antes de aplicar algoritmos baseados em distância como agrupamento k-means para garantir que todas as variáveis contribuam igualmente para a análise.
2. The process of transforming features to a common scale, typically between 0 and 1 or with mean 0 and standard deviation 1
Without feature scaling, a variable with values in the thousands would dominate the model over a variable with values between 0 and 1.
Sem normalização de características, uma variável com valores em milhares dominaria o modelo sobre uma variável com valores entre 0 e 1.
Feature scaling is a fundamental concept in machine learning and data science education in both Brazil and the USA. In Brazilian universities and tech companies, the Portuguese term 'normalização de características' is widely adopted and appears in technical documentation, while maintaining the English term 'feature scaling' in code comments and international collaborative projects. This reflects the global nature of machine learning research and development.
Look up more words on Fala2Me
The free English-Portuguese dictionary with real Brazilian accents, NYC slang, conjugator and more
Open Fala2Me →