MICROIMAGE PROCESSING OPTIMIZATION BASED ON RECURRING NEURAL NETWORK TRAINING AND IMPLICATIVE INFORMATIVE FEATURE SELECTION

In the article, the issue of intelligence analysis system development for identification and classification of indices of staff in higher educational institutions based on recurrent neural network (RNN) is stated. The methodology for modification of RNN computational schemes aimed at identification of random timing series is suggested. The methods and algorithms for data processing optimization are developed basing on informative feature matrix. The analysis of results on identification algorithms implementation is provided.