NEDA6 RDR 2019
Western Visayas Regional Development Report 2019
NEDA6 RDR 2020
Western Visayas Regional Development Report 2020
PN 2022-02
How Can DSWD's Social Pension Program for Indigent Senior Citizens be Improved?
RPS 2022-01
How Does the Philippines Fare in Meeting the ASEAN Economic Community Vision 2025?

PIDS WB 2021-1201
Analyzing the President's Budget for 2022
WASHaLOT 3.0: Production Process
WASHaLOT 3.0 Mass Production
Minimum Requirement Guidelines on WASH in Schools
Publication Detail
DP 2018-33: Exploring Neural Network Models in Understanding Bilateral Trade in APEC: A Review of History and Concepts

This paper seeks to understand certain frameworks that can be used to improve the analysis and prediction of trade flows within the Asia-Pacific Economic Cooperation economies using neural networks. Discussions include the history of neural network development, the biological neuron, the artificial neuron, and the potential use of neural networks in trade analysis. This paper also compares the different estimation procedures of the gravity model–-Ordinary Least Squares, Poisson Pseudo Maximum Likelihood, and Gamma Pseudo Maximum Likelihood-–with the neural network. Study finds that the neural network estimation of the gravity equation is superior over the other procedures in terms of explaining the variability of the dependent variable (export) around its mean and in terms of the accuracy of predictions.

Philippine Institute for Development Studies
Authors Keywords
Quimba, Francis Mark A.; Barral, Mark Anthony A.; artificial intelligence; neural networks; machine learning; gravity model; Ordinary Least Squares; Poisson Pseudo Maximum Likelihood; Gamma Pseudo Maximum Likelihood; trade, Asia-Pacific Economic Cooperation;
Download PDF Number of Downloads
Published in 2018 and available for Downloaded 467 times since December 26, 2018
Please let us know your reason for downloading this publication. May we also ask you to provide additional information that will help us serve you better? Rest assured that your answers will not be shared with any outside parties. It will take you only two minutes to complete the survey. You will answer the profile questions only once as long as you enter the same email address. Thank you.

To use as reference:
If others, (Please specify):
Name: (optional)
Email: (required, but will not display; please use the same email address when downloading another publication so that the profile questions will not appear)
If Prefer to self-describe, please specify:
Level of Education:
If employed either part-time or full-time, name of office:
If others, (Please specify):
Would you like to receive the SERP-P UPDATES e-newsletter? Yes No
Use the space below if you have any comment about this publication or SERP-P knowledge resources in general.