unsupervised pre-training
1 episode mention this concept
Knowledge Graph
Related concepts — line thickness indicates connection strength. Click any node to explore.
Related concepts
Concepts that appear alongside unsupervised pre-training across episodes.
- backpropagation
- l2 regularization (weight decay)
- input vector
- classification
- flow graph
- weights and biases
- vanishing gradient problem
- pre-activation
- empirical risk minimization
- dropout
- automatic differentiation
- loss function
- hyperbolic tangent (tanh) activation function
- sigmoid activation function
- stochastic gradient descent (sgd)
- rectified linear activation function (relu)
- activation function
- batch normalization
- softmax activation function
- feedforward neural networks