Python для обробки даних і моделювання фінансово-економічних показників

Abstract

The article considers aspects of efficient data processing. Much attention is paid to the problems that arise in modeling and fore-casting data and the role of research in decision-making. The stages of work with data and features that are inherent in each stage are deter-mined. A special place in the work is described by the possibilities of software data processing using Python, which is becoming increasingly popular due to simplicity, flexibility, open source, ease of working with data in various formats, as well as many developed packages that facilitate fast and efficient information processing. NumPy, Pandas, which provide data structures and functions that make working with structured data simple and fast, the most popular tool for data visualization Matplotlib, packages for various computational tasks SciPy, Statsmodels, as well as a package focused on machine learning Scikit-learn. An example of using Python for customs tasks is given. The authors have created a program for calculation, which uses the above packages. Regression models are being built to analyze the replenish-ment of the state budget of Ukraine with revenues from customs authorities at the expense of import and export duties. The analysis of mod-els on the basis of econometric methods of modeling is carried out and forecast estimates of receipts are calculated.