7. Neural Network

The computational graph or the neural network similar to many deep learning frameworks is our chosen glue/ executor of individual neurons or nodes. Assuming all nodes follow a predictable format usually using our Node abstraction, we can traverse the graph to execute nodes, or even use the longest path to calculate the minimum encryption parameters necessary for a leveled homomorphic cyphertext to successfully reach the end before the noise limit has been reached.

Computational Graph and Basic Components