This article discusses the application of deep learning algorithms to specific problems and explores the methods used in these processes. The presented research is based on the notion that "deep learning" is relevant to the machine learning community, and the research work of scientists is treated accordingly.

The interconnection of deep learning neural networks has been investigated. And machine learning sets can be conceptualized as deep learning sets and artificial intelligence sets are subsets. It also claims that the use of Convolutional Neural Networks (CNN) is an efficient and modern method for performing large-scale tasks in deep learning, especially image classification.

Examples of constructed tasks for predictive learning and analyzes the results of implementing a deep learning model. 11 000 documents were scanned and classified by title using Python. The documents have been converted to a regular model using the Tensorflow library and Keras. As a result, the data was collected and analyzed.