You can calculate everything with machine learning as long as you know a relevant truth. Just to do experiments that the matrix function model then has to adapt to. Assume all

input data can affect the virus. Then you insert a truth in the process that becomes label = 1 and without label = 0. Then train the model so that it grasps when something has been

inserted. It corresponds to 0.0.93,1.22, -0.01,1.01,1,0.1,0 from random mix of data. Remove the manual deposit and see what the model classifies for similar problems.

Idea – You can measure magnetic field noise and the time it takes for the virus to die in a hot water saline solution and in a hot water solution. It will be label 1 for saline and

label 0 for hot water only. Collect a lot of data X_train and y_train = 0,0,0, … 1,1,1 … which is a label truth> 1 min. Randomly mix data and then practice the ML function.

Then run the model on any substance in water to calculate the probability it takes before the virus dies.