non-deep neural network
[/nɑːn-dip ˈnjʊrəl ˈnɛtˌwɜːrk/]
nounpl: non-deep neural networks
rede neural rasa
1. A neural network with a limited number of hidden layers (typically one or two), used for simpler machine learning tasks as opposed to deep learning models
A non-deep neural network can be effective for linear classification problems that don't require multiple levels of abstraction.
Uma rede neural rasa pode ser eficaz para problemas de classificação linear que não exigem múltiplos níveis de abstração.
2. A machine learning model with shallow architecture, usually containing only input, one or two hidden layers, and output layers
The researchers chose a non-deep neural network to maintain computational efficiency on the embedded device.
Os pesquisadores escolheram uma rede neural rasa para manter a eficiência computacional no dispositivo embarcado.
This is primarily technical terminology used in artificial intelligence and machine learning communities worldwide. In Brazil, the term 'rede neural rasa' has gained prominence with the expansion of AI education and research institutions. The distinction between non-deep and deep neural networks became more culturally significant after the deep learning revolution of the 2010s, making this term important for differentiating older machine learning approaches from modern deep learning methods.
Look up more words on Fala2Me
The free English-Portuguese dictionary with real Brazilian accents, NYC slang, conjugator and more
Open Fala2Me →