Raqamli texnologiyalarning Yangi O‘zbekiston rivojiga ta’siri

 
RAQAMLI TEXNOLOGIYALARNING 
YANGI 
O‘ZBEKISTON
 RIVOJIGA 
TA’SIRI
 
Xalqaro ilmiy-amaliy konferensiyasi
 
social indicators into their models, enhancing the accuracy of GDP predictions. In this article, we 
will delve deeper into the significance of GDP and real GDP growth as key factors in assessing the 
health of an economy.
In their study, Richardson et al. aimed to improve nowcasts of real GDP growth in New 
Zealand by utilizing machine learning algorithms. They trained various popular ML algorithms 
using a large real-time dataset comprising approximately 550 New Zealand and international 
macroeconomic indicators. The authors compared the predictive accuracy of these nowcasts with 
several benchmarks, including autoregressive models, factor models, a large Bayesian VAR, and 
statistical models used at the Reserve Bank of New Zealand. Their findings indicated that machine 
learning algorithms outperformed the statistical benchmarks. Moreover, combining the nowcasts 
from different ML models further enhanced the overall performance, highlighting the gains in 
nowcasting accuracy achieved through the use of machine learning methods. [Adam 
Richardson,(2018)]
Gharte et al. investigated the use of machine learning techniques to improve the accuracy of 
GDP prediction. Analyzing various social, economic, and cultural parameters from 1970 to 2018, the 
authors built supervised learning models and compared the performance of three algorithms: 
Gradient Boosting, Random Forest, and Linear Regression. The study found that Gradient Boosting 
achieved the best prediction performance, followed by Random Forest and Linear Regression. The 
authors also developed a web application that estimated and forecasted the GDP of a country based 
on input attributes, highlighting the potential of machine learning techniques in GDP analysis and 
prediction. [Tanvi Gharte, (2022)] 
Maccarrone et al. compared different models for forecasting the real U.S. GDP. Using 
quarterly data from 1976 to 2020, they found that the machine learning K-Nearest Neighbour 
(KNN) model outperformed traditional time series analysis in capturing the self-predictive ability 
of the U.S. GDP. The authors explored the inclusion of predictors such as the yield curve, its latent 
factors, and macroeconomic variables to enhance forecasting accuracy, observing improved 
predictions for longer forecast horizons. The study highlighted the additional guidance provided by 
machine learning algorithms for data-driven decision-making. [Giovanni Maccarrone, (2021)] 
Wang et al. focused on the application of emotion recognition algorithms in analyzing and 
predicting financial market trends and economic growth. They highlighted the complexity of the 
financial market and economic growth as a highly intricate system and emphasized the need for 
accurate prediction results. The authors provided a detailed overview of existing financial 
development and economic growth forecasting issues, along with an introduction to emotion 
recognition algorithms. They delved into statistical emotion recognition methods, mixed emotion 
recognition methods, and emotion recognition methods based on knowledge technology. The study 

---

211