AI GlossaryDropoutENTRY — Deep Learning
[Deep Learning]
Dropout.
A regularization technique that randomly drops neurons during training to prevent overfitting.
In-depth explanation
01During each training step, dropout randomly sets a fraction of neuron outputs to zero. This prevents neurons from co-adapting too much and forces the network to learn more robust features. At inference time, all neurons are used but outputs are scaled. Dropout rate typically ranges from 0.1 to 0.5.
Examples
02EX. 01
Dropout rate of 0.5 in dense layers
EX. 02
Spatial dropout in CNNs
More in Deep Learning
0301Attention MechanismA technique that allows models to focus on relevant parts of the input when producing output.02Convolutional Neural Network (CNN)A neural network architecture designed for processing grid-like data such as images.03Fine-TuningAdapting a pre-trained model to a new task by training on task-specific data.04LSTM (Long Short-Term Memory)An RNN variant with gates that control information flow, enabling learning of long-term dependencies.05Recurrent Neural Network (RNN)A neural network architecture designed for sequential data with connections between nodes forming cycles.06Transfer LearningUsing knowledge learned from one task to improve performance on a different but related task.
[NEXT] — APPLY THE CONCEPT
Master Dropout.
Learn how to apply this concept with hands-on projects in our comprehensive AI programs.