DEVELOPMENT OF A VIRTUAL LABORATORY FOR CONVOLUTIONAL NEURAL NETWORKS

The article examines the potential for modernising the process of learning and research in the field of convolutional neural networks (CNN). The authors describe the architecture and functional capabilities of a laboratory that enables experimentation with various CNN parameters, significantly simplifying the learning and mastery of this technology. The virtual environment provides students with the opportunity to work with real data, visualise and analyse results without the need for expensive equipment. The laboratory also offers integration with databases, making it possible to store results in the system for future use.