Across
- 4. Function that introduces non-linearity in neural networks
- 5. Activation function with output between 0 and 1
- 7. Model inspired by human brain structure
- 8. Final layer producing prediction
- 9. First layer that receives raw data
- 10. Intermediate layer between input and output
- 13. Controls how much weights are updated
- 15. Error between predicted and actual values
Down
- 1. Algorithm used to update weights during training
- 2. Collection of data used for training
- 3. Connection strength between neurons
- 6. Model performs well on training but poorly on test data
- 11. Activation function that outputs zero for negative values
- 12. Additional parameter added to neuron output
- 14. One full pass through training dataset