Context, layers and average best linear mapping. I…
0 reactions
Context, layers and average best linear mapping.
I honestly don't mean to wipe trillions off the stock market but you can view a neural network layer as context choosing a linear mapping. Then BP comes along and makes that the best average linear mapping for that layer and context.
For a particular input a neural network is a composition of best average linear mappings.
https://archive.org/details/best-average-linear-mapping
https://archive.org/details/a-compositional-view-of-deep-networks-as-fractional-associative-memories
#technology
I honestly don't mean to wipe trillions off the stock market but you can view a neural network layer as context choosing a linear mapping. Then BP comes along and makes that the best average linear mapping for that layer and context.
For a particular input a neural network is a composition of best average linear mappings.
https://archive.org/details/best-average-linear-mapping
https://archive.org/details/a-compositional-view-of-deep-networks-as-fractional-associative-memories
#technology
| 4,000 mlx | |
| 4 | |
| 0 |