sanguovbobo / trade-analysis Goto Github PK

Analysing the International Trade using Extended Gravity Model, Neural Networks and Random Forest. We set to Analyse tradeflow patterns using both Intuitive and Extended Gravity Model. Further Implemented other Gravity Model Estimators like Pseudo Maximum Likelihood Estimator (PPML) And GPML. Implementation of Neural Network and Random Forest on Data from 1948-2019 containing 18 Countries and Comparison of predictive powers using R2. The Results Signify that Random Forest is better than Neural Network, Fixed Estimation, PPML and OLS. Analysing Trade Pattern between India and USA between 2014-2019 estimated using Random Forest ending the Study by suggesting the Future Research Areas,