Minimizing the Risk of Data Breaches in U.S. Commercial Entities Using Predictive Analytics
Open Access DepositedDownloadable Content
both Logistic Regression alone and the Logistic Regression-SVM hybrid were second, followed by RF, SVM, and LSTM.
LSTM, the Linear Regression-SVM hybrid, the RF-LSTM hybrid, RF, and SVM. Finally, the RF-LSTM hybrid model achieved the best performance in predicting the occurrence of ransomware incidents, followed by the other models, listed in order of performance
Minimizing the Risk of Data Breaches in U.S. Commercial Entities Using Predictive Analytics Data breaches and ransomware attacks on U.S. commercial entities are happening more often and have become more complex. The capability to predict data breaches, the number of records breached, and ransomware attacks will contribute to more robust and proactive cybersecurity practices that U.S. commercial entities need to combat this problem. This research aims to evaluate predictive models for accuracy. The data from eleven sources, which ranged from as few as 10 records to as many as 35K records, were gathered for this research. Ultimately, only the data from the sources, Privacy Rights Clearinghouse, Kaggle, and The University of Queensland, which provided the three largest datasets, was used. Models based on long short-term memory (LSTM), regression analysis, random forest (RF), and support vector machine (SVM), in addition to two hybrid models comprised of regression analysis and SVM, and RF and LSTM. This research compared the predictive models and determined which models have the highest accuracy.After the model comparisons, this research found that the Linear Regression-SVM hybrid model achieved the best performance in predicting the annual number of data breaches, followed by the other models in order of performance
LSTM, Linear Regression, RF, the RF-LSTM hybrid, and SVM. Also, the Linear Regression model achieved the best performance in predicting the annual number of records breached, followed by the other models in order of performance
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.