article · Computational Urban Science
Accurate economic forecasting relies heavily on measuring Gross Domestic Product to evaluate national and global economic performance. A new predictive architecture combines Pearson correlation with a Long Short-Term Memory Recurrent Neural Network to forecast output within urban profiling areas. Pearson correlation isolates the most significant features linked to the target variable. Data preparation includes median imputation and data normalisation. The method was evaluated using two datasets: one comprising 227 instances across 20 features, and another containing 61 instances representing historical Indian growth data between 1961 and 2021. A parameter transfer strategy fine-tunes parameters learned from the first dataset onto the second to enhance predictive accuracy. Benchmarking against multiple error metrics and the coefficient of determination showed that the framework outperformed competing regression models, reaching a determination coefficient of 99.99 percent with minimal error rates.
Gross Domestic Product provides essential information about the scale and operational health of an economy. Enhancing predictive algorithms allows analysts to forecast economic growth trajectories more precisely. Demonstrating near-perfect correlation on benchmark datasets indicates that integrating feature selection with recurrent neural networks and transfer learning offers a powerful computational route for macroeconomic forecasting.
The model could enable automated economic forecasting software for macroeconomic analysts, urban planning agencies, and policy organisations. The abstract demonstrates an applied and tested stage of algorithmic research validated on historical data, though it indicates no immediate commercial product, deployment infrastructure, or commercialisation pathway.
AI-generated from the published abstract. Always read the original work before citing.
Abstract Gross Domestic Product (GDP) is significant for measuring the strength of national and global economies in urban profiling areas. GDP is significant because it provides information on the size and performance of an economy. The real GDP growth rate is frequently used to indicate the economy’s health. This paper proposes a new model called Pearson Correlation-Long Short-Term Memory-Recurrent Neural Network (PC-LSTM-RNN) for predicting GDP in urban profiling areas. Pearson correlation is used to select the important features strongly correlated with the target feature. This study employs two separate datasets, denoted as Dataset A and Dataset B. Dataset A comprises 227 instances and 20 features, with 70% utilized for training and 30% for testing purposes. On the other hand, Dataset B consists of 61 instances and 4 features, encompassing historical GDP growth data for India from 1961 to 2021. To enhance GDP prediction performance, we implement a parameter transfer approach, fine-tuning the parameters learned from Dataset A on Dataset B. Moreover, in this study, a preprocessing stage that includes median imputation and data normalization is performed. Mean Square Error, Mean Absolute Error, Root Mean Square Error, Mean Absolute Percentage Error, Median Absolute Error, and determination coefficient (R 2 ) evaluation metrics are utilized in this study to demonstrate the performance of the proposed model. The experimental results demonstrated that the proposed model gave better results than other regression models used in this study. Also, the results show that the proposed model achieved the highest results for R 2 , with 99.99%. This paper addresses a critical research gap in the domain of GDP prediction through artificial intelligence (AI) algorithms. While acknowledging the widespread application of such algorithms in forecasting GDP, the proposed model introduces distinctive advantages over existing approaches. Using PC-LSTM-RNN which achieves high R 2 with minimum error rates.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1007/s43762-024-00116-2
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
