Scholarlink

<p>This research attempts to highlight the strategies that need to be implemented by countries to escape the middle-income trap by sequencing the predictors responsible for economic transition from middle income to high income category using machine learning techniques. Random Forest (RF) and Support Vector Machine (SVM) are the two modelling techniques utilized on data of 85 nations’ covering 28 variables. A mixed approach that combines Principal Component Analysis (PCA) is used to address multicollinearity while preserving explained variance. The identified factors encompass the interplay of structural, macroeconomic, institutional indicators within a given demographic dynamics that contour the progression of countries from middle to high-income groupings. The PCA-transformed determinants are trained via RF and SVM. Subsequently, the original variables are recovered and analysed through back mapping. The study also uses Logistic Regression (LR) as a simple linear baseline model for comparison of Variable Importance Plots (VIPs) obtained from the RF and SVM techniques. The prioritization of industrial upgradation, sophistication of exports, human capital development and improvement in governance emerge out to be unique set of contributors to transition. This strengthens the view that economic transitions are not determined by isolated interventions but by the synergistic effects of macroeconomic, institutional, and structural dynamics. Future research may search for dynamic connections among policy instruments, feedback mechanisms in trade and institutional ecosystems. The resilience of transition pathways in the face of external shocks is equally pertinent.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Analytical insights on economic transition: ranking key determinants using machine learning approach

  • 期刊论文

  • 发表日期：

    2026年2月25日

查看全文
  • A. Jiran Meitei,
  • Pratibha Rai,
  • Vijeta Pundir,
  • Priya Gupta

摘要

This research attempts to highlight the strategies that need to be implemented by countries to escape the middle-income trap by sequencing the predictors responsible for economic transition from middle income to high income category using machine learning techniques. Random Forest (RF) and Support Vector Machine (SVM) are the two modelling techniques utilized on data of 85 nations’ covering 28 variables. A mixed approach that combines Principal Component Analysis (PCA) is used to address multicollinearity while preserving explained variance. The identified factors encompass the interplay of structural, macroeconomic, institutional indicators within a given demographic dynamics that contour the progression of countries from middle to high-income groupings. The PCA-transformed determinants are trained via RF and SVM. Subsequently, the original variables are recovered and analysed through back mapping. The study also uses Logistic Regression (LR) as a simple linear baseline model for comparison of Variable Importance Plots (VIPs) obtained from the RF and SVM techniques. The prioritization of industrial upgradation, sophistication of exports, human capital development and improvement in governance emerge out to be unique set of contributors to transition. This strengthens the view that economic transitions are not determined by isolated interventions but by the synergistic effects of macroeconomic, institutional, and structural dynamics. Future research may search for dynamic connections among policy instruments, feedback mechanisms in trade and institutional ecosystems. The resilience of transition pathways in the face of external shocks is equally pertinent.