ECONOMETRIC ANALYSIS OF AZERBAIJAN'S GDP USING SINGULAR SPECTRAL ANALYSIS (SSA)

---

Əhmədova E.

International Journal of Sciences and Research, pp.225-234, 2020 (Scopus)

• Nəşrin Növü: Article / Article
• Nəşr tarixi: 2020
• jurnalın adı: International Journal of Sciences and Research
• Jurnalın baxıldığı indekslər: Scopus
• Səhifə sayı: pp.225-234
• Açıq Arxiv Kolleksiyası: Rəqəmsal Yaddaş Kolleksiyası
• Adres: Bəli

Qısa məlumat

Conclusion

Based on the data analysis, the three time periods (1997-2004, 2005-2014 and 2015-

2018) that combine various trends and indicator variation ranges have been identified. Each

period involved models that considered trends only. The designed models proved to be

adequate (r2=68-80%). However, additional introduction of the seasonal component to the

model improved it: r2 = 78-84%, which is 2-10% more than for the model that takes into

account only trends. Nevertheless, the Fourier analysis suggests that there is an additional cyclic

component of 43-44 months, and its introduction into the model will allow to achieve even

better outcomes. Additional introduction of a cyclic component of 43-44 months into the model

has increased the coefficient of determination by 1-3%