Web service for detecting fraudulent transactions using artificial intelligence technologies

Abstract

These theses describe the stages of software implementation of a software system for detecting fraudulent transactions using machine learning and artificial intelligence technologies. The description of the stages of software implementation of this system includes: analysis of similar software systems, design of the software system architecture, comparison of machine learning algorithms and selection of the most effective or effective, training and testing of the model, selection of a technology stack for software implementation, as well as the main steps of software implementation. The stage "analysis of similar software systems" contains a detailed description of the metrics, characteristics and features that should be paid attention to when implementing a similar system. This stage helps to identify the strengths and weaknesses of similar systems. The stage "design of the software system architecture" contains a description of the design of the software system architecture with a justification of the effectiveness of this architecture. The stage "comparison of machine learning algorithms and selection of the most effective or effective" contains a description of existing metrics of the effectiveness of machine learning models, as well as the most important metrics of effectiveness specifically for those models that are designed to detect fraudulent transactions. The stage "model training and testing" contains a description of the preparation of the dataset for model training, the design of the machine learning model, the training of the model, and its testing. The stage "technology stack selection for software implementation" contains a description of the selection of a technology stack for software implementation of a software system for detecting fraudulent transactions with a justification for why this particular technology stack was chosen. The stage "basic steps of software implementation" contains a description of the basic steps of software implementation of such a software system. Design and software implementation of a software system for detecting fraudulent transactions is relevant today, since the problem of transaction fraud and the adaptation of fraudsters to existing software systems designed to combat transaction fraud is relevant. The purpose of writing these theses is to describe the stages of software implementation of a system for detecting fraudulent transactions using machine learning and artificial intelligence technologies, which would compare the strengths of existing similar systems and minimize their shortcomings. Also, the theses contain a description of the preparatory actions without which the development of a software system for detecting fraudulent transactions using machine learning and artificial intelligence technologies is impossible.