Web Platform for Training and Applying Machine Learning Models and Neural Networks

The thesis on topic "Web platform for training and applying machine learning and neural network models" contains 125 pages, 48 figures and 42 sources. In today's world, artificial intelligence (AI) plays a key role in business development. However, implementing AI can be a complicated task, especially for companies with no experience in the field. An automated machine learning platform that makes AI accessible to a wider range of companies, speeds up the implementation process, and reduces software development costs can be a solution to this problem. Given the shortage of qualified ML specialists, using platforms to train neural networks without the need to write program code is of particular importance. This approach enables rapid creation of machine learning models and automates data processing and visualization of results. This paper proposes a solution to the above problems in the form of developing a web-based platform for training and applying machine learning models and neural networks. The developed platform provides solutions to problems related to three main areas of machine learning: classical machine learning, computer vision and natural language processing.