Analysis of the use of artificial intelligence algorithms for deep analysis of financial data

DOI: 10.31673/2409-7292.2024.030005

Abstract

It is difficult to imagine the modern world without the use of artificial intelligence. This seemingly futuristic technology has already become an integral part of our reality. Neural networks are so accessible that they can be used not only by large corporations like Microsoft or Google, but also by ordinary users in their daily activities. Artificial intelligence is now being used to teach, assist in work, and perform boring and routine tasks. That is, such a tool can support the user in any activity, even in the forecasting and analysis of his financial situation both in the present and in the future. The paper analyzes the potential possibility of using different types of neural models to solve problems related to work with personal finances. Artificial intelligence systems are capable of processing large amounts of data, identifying patterns and making fairly accurate predictions that will allow users to make informed decisions about their finances. This approach will provide an opportunity to analyze costs, revenues and other financial aspects, helping to optimize budgets, foresee potential risks and opportunities for growth. This article examines artificial intelligence models and algorithms capable of analyzing financial data and providing analytics based on it. In particular, neural networks such as recurrent neural networks RNNs and long short-term memory LSTMs are considered, which efficiently process time series of financial indicators. Generative-competitive GANs are used to generate synthetic data and detect anomalies. Machine learning methods, including regression, decision trees, random forest and gradient boosting, that allow forecasting and classification of financial indicators and metrics. Also, clustering algorithms such as k-means and DBSCAN, which can help in customer segmentation and anomaly detection, will be considered. Natural language processing models such as Llama 3 8B and GPT-3 to analyze users' text financial data and help generate insights.

Keywords: artificial intelligence, forecasting, financial analysis, budget optimization, machine learning, regression, decision trees, random forests, clustering, natural language processing, Llama 3 8B.