06-09-2023, 05:20 PM
If I understand correctly, you are trying to use some neural networks to find a script which would be visually most similar to Voynich script.
Sorry but that won't help with deciphering the manuscript. Imagine a cipher where standard set of letters A..Z is replaced by the same letters
A..Z according to some rules (substitution cipher, transposition cipher, nomenclator etc.)
In that case both scripts are immediately known, yet it doesn't make you closer to get the hidden meaning.
The neural network will probably find "something" if you feed it enough data from different parts of the world. Some ancient or modern script
from India or Nepal or Thailand. Which won't make sense in the historical context of Voynich Manuscript.
Sorry but that won't help with deciphering the manuscript. Imagine a cipher where standard set of letters A..Z is replaced by the same letters
A..Z according to some rules (substitution cipher, transposition cipher, nomenclator etc.)
In that case both scripts are immediately known, yet it doesn't make you closer to get the hidden meaning.
The neural network will probably find "something" if you feed it enough data from different parts of the world. Some ancient or modern script
from India or Nepal or Thailand. Which won't make sense in the historical context of Voynich Manuscript.