Deep Neural Networks

Recall, Optimization, Analysis of the different methods, Activation functions, Various ReLUs, Softmax, Error functions, Cross-entropy, Negative log-likelihood, Regularization, Batch normalization, Weight regularization

It is a method used for optimizing network parameters. It involves finding the gradient of the function and updating the parameters in the opposite direction of the gradient.

Momentum methods, Adaptive Gradient methods, ADAM

<p>No, reaching the global minimum can lead to overfitting and loss of generalization capabilities.</p> Signup and view all the answers

<p>Overfitting occurs when a model learns the training data too well and loses its ability to generalize to new, unseen data.</p> Signup and view all the answers