Forecasting the Sector wise Gross Domestic Product of India

Loganathan, Ramanan (2023) Forecasting the Sector wise Gross Domestic Product of India. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration manual) Download (2MB) | Preview

Abstract

This analysis delves into the significance of temporal sequence prediction techniques for forecasting India’s Gross Domestic Product (GDP), a pivotal economic metric. The dataset encompasses historical GDP values across eight distinct sectors and GDP at factor cost, presenting a comprehensive overview of the economy. The primary objective is to employ and assess diverse forecasting approaches to accurately predict GDP trends. The initial phase involves meticulous data preprocessing, including managing missing values, scaling data, and partitioning it into training and testing subsets. Four distinct methodologies, specifically SARIMA (Seasonal Autoregressive Integrated ‘Moving Average), ARIMA (Autoregressive Integrated Moving Average), RNN (Recurrent Neural Network), and LSTM (Long Short-Term Memory), were utilized. The ‘statsmodels’ library is utilized for SARIMA and ARIMA, while ‘keras’ facilitates the implementation of RNN and LSTM models. The training process involves constructing SARIMA and ARIMA models on the training set, followed by forecasting GDP values for the test set. Performance metrics such as Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Squared Error (MSE) evaluate the precision of the predictions. For RNN and LSTM, a sequence-based training approach is adopted. The findings of this examination underscore the efficacy of the models in predicting GDP values across sectors. Notably, the LSTM model demonstrates superior precision, boasting the lowest RMSE among the models. The implications of this analysis are substantial, offering valuable insights for decision-makers in various economic sectors. The robustness of these forecasting models assists in generating well-informed projections, thereby contributing to strategic deliberations and policy formulation. Essentially, this research illuminates the pivotal role of accurate GDP prediction and advocates for the strategic integration of suitable forecasting models based on data attributes. The outcomes pave the path for future advancements in economic forecasting, highlighting the ever-increasing significance of data-driven insights in shaping economic paths.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Rustam, Furqan UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HB Economic Theory > Business Cycles. Economic Fluctuations H Social Sciences > Economics Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	29 Nov 2024 13:02
Last Modified:	29 Nov 2024 13:02
URI:	https://norma.ncirl.ie/id/eprint/7210

Actions (login required)

View Item