Predicting of device parameters of of N,N'-Dioctyl3,4,9,10-perylenedicarboximide thin Film based photodiode: Machine learning approach