arbiter
JournalsAccount

Merger and acquisition prediction based on deep learning with attention mechanism

Liangyong Wan et al.

China Journal of Accounting Research2026https://doi.org/10.1016/j.cjar.2025.100459article
AJG 2ABDC A
Weight
0.50

Abstract

This study proposes a novel attention-based deep neural network (AttDNN) model specifically designed for predicting mergers and acquisitions (M&A). The model extends existing deep learning frameworks by incorporating M&A-specific features, regularization layers, and an attention mechanism that emulates human cognitive processes to structure M&A drivers and improve predictive performance. Empirical results demonstrate that the AttDNN model significantly outperforms traditional algorithms in forecasting M&A outcomes, achieving a 29.2 % improvement in predictive accuracy over conventional deep learning methods. This study provides valuable insights at the intersection of artificial intelligence and financial economics, offering practical implications for financial strategy and corporate M&A decision-making.

Open via your library →

Cite this paper

https://doi.org/https://doi.org/10.1016/j.cjar.2025.100459

Or copy a formatted citation

@article{liangyong2026,
  title        = {{Merger and acquisition prediction based on deep learning with attention mechanism}},
  author       = {Liangyong Wan et al.},
  journal      = {China Journal of Accounting Research},
  year         = {2026},
  doi          = {https://doi.org/https://doi.org/10.1016/j.cjar.2025.100459},
}

Paste directly into BibTeX, Zotero, or your reference manager.

Flag this paper

Merger and acquisition prediction based on deep learning with attention mechanism

Flags are reviewed by the Arbiter methodology team within 5 business days.

Evidence weight

0.50

Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40

F · citation impact0.50 × 0.4 = 0.20
M · momentum0.50 × 0.15 = 0.07
V · venue signal0.50 × 0.05 = 0.03
R · text relevance †0.50 × 0.4 = 0.20

† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.