,
Research Scholar, Faculty of Management, SRM Institute of Science and Technology, Kattankulathur, Chengalpattu, Tamil Nadu, India
Assistant Professor, Faculty of Management, SRM Institute of Science and Technology, Kattankulathur, Chengalpattu, Tamil Nadu, India
The decisions made in investing among salaried individuals are guided by financial objectives, risk appetite, and demographics, with gold and real estate continuing to be some of the most favored ones in India. In this study, investment preference for gold and real estate among salaried individuals in India and a prediction model for predicting preferred investment types based on demographics are explored. Data was obtained using a questionnaire that was filled out online through Google Forms by 300 salaried individuals from different sectors. Label correlation guided borderline oversampling (LCGBO) was used to solve class imbalance, and an explicit feature interaction-aware graph neural network (EFIGNN) was
developed to predict investment preference. One-way ANOVA, T-test, and Chi-square were used to determine the influence of age, income level, and gender on investment choice. According to ANOVA results, there was a significant age effect on the preference for real estate (p = 0.012) and gold (p = 0.021). On the other hand, the Chi-square showed a significant relationship between income and investment preference (χ² = 8.13 for gold and χ² = 2.95 for real estate). T-test showed a significant difference in gender (p = 0.001 for real estate and p = 0.005 for gold). EFIGNN achieved 98% accuracy and an RMSE of 0.1%, outperforming ANN, LSTM, and GNN. Real estate emerged as the most preferred investment,with demographic factors significantly shaping investment behavior and EFIGNN demonstrating strong predictive performance for investment preference modeling.
This is an open access article distributed under the Creative Commons Attribution Non-Commercial License (CC BY-NC) License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
0
The statements, opinions and data contained in the journal are solely those of the individual authors and contributors and not of the publisher and the editor(s). We stay neutral with regard to jurisdictional claims in published maps and institutional affiliations.