A Hybrid ANN-SVM Framework for Ransomware Detection with Imbalanced Class Consideration

Ransomware puts businesses at danger of financial, reputational, and data loss. Strong crisis response processes and cybersecurity infrastructure are necessary to halt these types of assaults. Integrating ML and DL technology may fortify your viral defenses. Combining an Artificial Neural Network (ANN) with a Support Vector Machine (SVM) was a novel approach to infection detection in this study. While ANN excels in identifying complex data structures, SVM excels at dealing with multi-dimensional data. Together, we can improve malware detection, which strengthens our defenses against emerging threats. During the testing phase, a dataset from the CIC institute was utilized along with hyperparameters such as loss, accuracy, val-accuracy, and val-loss. Finding out how well the machine learning model can detect malware relies on these signals. You can see how well the model performed on the data from the training set by looking at the loss and accuracy metrics. Metrics like val-accuracy and val-loss reveal how much the system has improved at being consistent and reliable in actual use. There is hope that businesses may increase their security by implementing ML and DL. To better protect against ever-changing cyber threats, it might be useful to implement a thorough method for detecting ransomware that considers file attributes and activity patterns. Businesses may lessen their susceptibility to ransomware attacks, keep operations running, and safeguard vital information by using this unified approach. To aid companies in safeguarding their digital assets and remaining vigilant against new threats, effective cybersecurity plans utilize modern technology. By exploring and using new technology, a company may show that it is dedicated to keeping up with cybersecurity improvements and responding to a constantly evolving threat landscape.