Data Science and Neural Network in Cryptography

The use of neural networks for cryptography is demonstrated in this paper. There are two stages to the system. In the first stage, computer vision completely non numbers (NPRNGs) are created, and the outputs are checked for randomization utilizing random variation tests from the National Institute of Standards and Technology (NIST). Inside the second phase, NPRNGs are used to create a neural network-based cryptosystem. Data that has been encrypted using non-linear algorithms is subjected to decryption efforts using two identical artificial neural networks in just this crypto algorithm (ANNs). Non-linear encryption is modelled utilising apparent connection capabilities in the first neural network. The next neuralnetwork uses decision-making functionality to encode and decode data.