AN EVALUATION OF DOMESTIC AND CROSS-BORDER MERGERS & ACQUISITIONS IN INDIA USING DEEP LEARNING
Keywords:
Long Short-Term Memory, Krill Herd Algorithm, Merger and Acquisitions, Recurrent Neural Network.Abstract
In recent years, the landscape of mergers and acquisitions (M&A) has witnessed unprecedented shifts, with India emerging as a focal point for domestic and cross-border transactions. This study presents an in-depth analysis of domestic and cross-border M&A activities within the Indian economic landscape, leveraging the capabilities of deep learning methodologies. In this research, a Deep Learning framework named Long Short-Term MemoryRecurrent Neural Network (LSTM-RNN)was employed for analyzing the intricate patterns and dependencies in M&A data.The models are trained to predict M&A outcomes based on learned relationships between macroeconomic indicators and deal outcomes. Furthermore, a Krill Herd Algorithm was employed for optimizing the parameters of LSTM-RNN, reducing the loss function. The KHA fine-tunes the hyperparameters of LSTM-RNN and improves the prediction performances. The presented study was implemented in MATLAB for the publically available M&A dataset. The experimental analysis demonstrates that the developed method acquired a greater accuracy of 98.7% and a significantly lower RMSE of 0.124%. Furthermore, a comprehensive comparative study with the existing techniques manifests the efficiency of the proposed approach in examining the M&A deal outcomes for the macroeconomic factors.