On the approach to forecasting indicators of socio-economic development of the region based on indirect indicators

Economic and social development requires constant modernization of the regional management system based on the system of key socio-economic indicators of the region's development and methods of their analysis and forecasting. The article proposes a comprehensive approach to forecasting based on the application of classical forecasting methods for existing time series of statistical indicators and by identifying and analyzing indirect semantically close indicators to a new indicator in the absence of the necessary time series for forecasting. The article provides a general methodology for obtaining a forecast and describes in detail the method for constructing a forecast estimate of the change dynamics in the estimated indicator as well as a description of the AutoML library with open source FEDOT, which was used to build forecasts. The issue of constructing and optimizing a combined forecast with the aid of automatic machine learning tools is considered. At the end of the article, the result of an experiment on predicting the indicators “Population of the subject of the Russian Federation” and “Life expectancy at birth” according to the proposed approaches and a comparison of the findings is presented. It can be concluded that the suggested approach to making a predictive assessment of the change dynamics in the estimated indicator by identifying indirect indicators can be applied to socio-economic indicators of the development of the region.

1. Decree of the President of the Russian Federation dated February 4, 2021 No. 68 «On estimating the effectiveness of the activities of senior officials (heads of the highest executive bodies of state government) of the constituent entities of the Russian Federation and the activities of executive authorities of the constituent entities of the Russian Federation». (In Russ.)

2. Resolution of the Government of the Russian Federation dated April 3, 2021 No. 542 «On approval of methods for calculating indicators for evaluating the effectiveness of the activities of senior officials (heads of the highest executive bodies of state government) of the constituent entities of the Russian Federation and the activities of executive authorities of the constituent entities of the Russian Federation». (In Russ.)

3. Makridakis S., Spiliotis E. and Assimakopoulos V., Statistical and machine learning forecasting methods: Concerns and ways forward». PLoS ONE. 2018;13(3):e0194889.

4. Makridakis S., Hibon M. The M3-Competition: results, conclusions and implications. International jornal of forecasting. 2000;16(4):451–476.

5. Frenkel A., Volkova N., Surkov A. and Romanyuk E. The application of ridge regression methods when combining forecasts. Finance: Theory and Practice. 2018:124–133. (In Russ.)

6. Frenkel A., Volkova N., Surkov A. and Romanyuk E. Comparative analysis of methods for constructing a combined forecast. Voprosy statistiki. 2017:17–27. (In Russ.)

7. Frenkel A. and Surkov A. Determination of weighting factors in combining forecasts. Voprosy statistiki. 2017:17–27. (In Russ.)

8. Polonskaia I., Nikitin N., Revin I., Vychuzhanin P. and Kalyuzhnaya A. Multi-objective evolutionary design of composite data-driven models. 2021 IEEE Congress on Evolutionary Computation (CEC). 2021:926–933.

9. Xie Y. and Jiang H. Stock market forecasting based on text mining technology: A support vector machine method. Journal of Computers. 2017;12(6):500–510.

10. Kuh S., Chiu G. and Westveld A. Modeling national latent socioeconomic health and examination of policy effects via causal inference. 2019. Available by: https://arxiv.org/abs/1911.00512.

11. Yagi I., Masuda Y. and Mizuta T. Analysis of the impact of high-frequency trading on artificial market liquidity. IEEE Transactions on Computational Social Systems. 2020;7:1324–1334.

12. He Q., Panp P. and Si Y. Multi-source transfer learning with ensemble for financial time series forecasting. 2020 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology (WI-IAT). 2020;1:227–233.

13. Weeraddana D., Khoa N., O Neil L., Wang W. and Cai C. Energy Consumption Forecasting Using a Stacked Nonparametric Bayesian Approach. Machine Learning and Knowledge Discovery in Databases. Applied Data Science and Demo Track. ECML PKDD 2020. Lecture Notes in Computer Science. 2021;12461.

14. Rajapaksha D., Bergmeir C. and Hyndman R.J. LoMEF: A Framework to Produce Local Explanations for Global Model Time Series Forecasts. 2021. Available by: https://arxiv.org/abs/2111.07001.

15. Tilly S. and Livan G. Macroeconomic forecasting with statistically validated knowledge graphs. 2021. Available by: https://arxiv.org/abs/2104.10457.

16. Huang J., Chang K., Xiong J. and Hwu W. Open relation modeling: Learning to define relations between entities. 2021. Available by: https://arxiv.org/abs/2108.09241v1.

17. Nimishakavi M., Saini U. and Talukdar P. Relation schema induction using tensor factorization with side information. 2016 Conference on Empirical Methods in Natural Language Processing. 2016:414–423.

18. Feurer M., Klein A., Eggensperger K., Springenberg J.T., Blum M., and Hutter F. Auto-sklearn: efficient and robust automated machine learning. Automated Machine Learning. Springer, Cham. 2019:113–134.

19. Drori I., Krishnamurthy Y., Rampin R., Lourenco R., One J., Cho K., Silva C. and Freire J. AlphaD3M: Machine learning pipeline synthesis. ICML AutoML workshop. 2018.

20. Coors S., Schalk D., Bischl B. and Rügamer D. Automatic Componentwise Boosting: An Interpretable AutoML System. 2021. Available by: https://arxiv.org/abs/2109.05583.

21. Luo Z., He Z., Wang J., Dong M., Huang J., Chen M. and Zheng B. AutoSmart: An Efficient and Automatic Machine Learning framework for Temporal Relational Data. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining. 2021:3976–3984.