Set A-Deep Learning

4. This neural network type uses convolution operations and is popular in image recognition tasks.
6. A type of learning where the system infers patterns without labeled outcomes.
7. A visual representation extracted from audio to train music genre classifiers.
8. The function used to introduce non-linearity in a neural network, commonly ReLU or sigmoid.
9. An unsupervised model that learns efficient representations by encoding and decoding input data.
10. A type of neural network suitable for sequential data like text or time series.

1. A method that involves normalizing input values across a batch during training.
2. The phenomenon where a model performs well on training data but poorly on new data.
3. The method by which neural networks are trained, adjusting weights to reduce error.
5. This mechanism in RNNs helps retain long-term dependencies.