Scopus
Machine learning-based Covid-19 forecasting: Impact on Pakistan stock exchange
International Journal of Agricultural and Statistical Sciences · Haziran 2021
Makale Bilgileri
DergiInternational Journal of Agricultural and Statistical Sciences
Yayın TarihiHaziran 2021
Cilt / Sayfa17 · 53-61
Scopus ID2-s2.0-85109077746
Özet
Machine learning methods have proved to be a prominent study field while solving composite real-world problems. Presently, the world is suffering from the Covid-19 pandemic disease, and its impact needs to be forecasted. The stock exchange is the backbone of any country's economy. After the Covid-19, the stock exchange was too affected. This study is based on the effect of Covid-19 on the Pakistan stock exchange. Pakistan's Covid-19 daily new cases were obtained from the website "our world in data" and stock exchange data KSE-100 from "Yahoo Finance." Machine learning techniques were used to forecast the stock exchange and Covid-19 daily new cases using a wave Ist dataset from 27th February 2020 to 2nd September 2020. Results prove that Pakistan's stock exchange KSE-100 index has shown a positive increase in stock returns. The accuracy of XGBoost is best as compared to the GLMNet method. These two forecasting methods were compared to different accuracy metrics. Best and suitable methods were selected on minimum MAE, MAPE, MASE, SMAPE, RMSE, and maximum R2value. These projections helped the government to make strategies for stock exchange KSE-100 and fight against a pandemic disease.
Yazarlar (6)
1
Iqra Sardar
2
Kadir Karakaya
3
Tatiana Makarovskikh
4
Mostafa Abotaleb
5
Syed Aflake
6
Pradeep Mishra
Anahtar Kelimeler
Covid-19
Machine learning
Stock exchange
Time series forecasting
Kurumlar
Jawaharlal Nehru Krishi Vishwa Vidyalaya
Jabalpur India
Riphah International University
Islamabad Pakistan
Selçuk Üniversitesi
Selçuklu Turkey
South Ural State University
Chelyabinsk Russian Federation
Metrikler
19
Atıf
6
Yazar
4
Anahtar Kelime