Integrating Artificial Intelligence with Cyber Threat Intelligence to Predict Financial Exposure in Data-Driven Enterprises
Year: 2025
Type: article
Abstract: As global reliance on digital infrastructure intensifies, the financial ramifications of cybersecurity incidents have become a critical concern for organizations worldwide.Quantifying and predicting these financial losses is essential for effective risk management, yet it remains a significant analytical challenge.This study addresses this problem by analyzing a decade of global cybersecurity threat data (2015)(2016)(2017)(2018)(2019)(2020)(2021)(2022)(2023)(2024) to model the financial impact of cyber attacks, evaluate the predictive power of various machine learning models, and identify underlying incident archetypes.Employing a multi-stage methodology, the research began with an exploratory data analysis, followed by a comparative evaluation of four regression models: Linear Regression, Random Forest, Gradient Boosting, and Support Vector Regression.Feature importance was extracted from the models, and K-Means clustering was used to derive a data-driven taxonomy of incidents.The study reveals a persistent and costly threat landscape with no significant decrease in resolution times over the decade.A crucial insight emerged from the predictive modeling: a simple Linear Regression model showed weak but positive predictive power.In contrast, more complex non-linear models failed, producing negative R² scores and indicating a propensity to overfit.The most significant predictors of financial loss were identified as the "Number of Affected Users" and "Incident Resolution Time."Finally, the clustering analysis successfully segmented incidents into distinct archetypes, such as "Mass Data Breach" and "High-Stakes Targeted Attack."This research concludes that noisy, linear relationships govern the financial ramifications PriMera Scientific Engineering https://primerascientific.com/psen Integrating Artificial Intelligence with Cyber Threat Intelligence to Predict Financial Exposure in Data-Driven Enterprises 14 This paper aims to make a substantive contribution to the cybersecurity and data science literature.By analyzing a comprehensive dataset spanning ten years, this research provides a longitudinal perspective that is often missing in contemporary studies.The pri-of cyber attacks and underscores the principle of model parsimony, providing a quantitative framework for understanding cyber risk and the primary factors that mitigate financial damage.
DOI: 10.56831/psen-07-214
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4414199598
FWCI:
Cites:
Cited by:
Related to: