Application of the neural network by the method of backpropagation of the error for a text data set

Makhnev Sergei Alexandrovich

The article discusses a method for processing textual, unstructured data sets for recognizing them graphically for a neural network using the backpropagation method. The purpose of this work is to develop a neural network and train it with data prepared by the semantic-graphic method. The task of the research is to conduct training and subsequent verification of the effectiveness of pattern recognition. Also, the article presents an analysis of the application of the processed array of input data, both with the use of ready-made tensorflow libraries, and in a properly developed neural network.

Keywords:neural network, backpropagation method, tensorflow.