Recent Papers

The proliferation of Android smartphones has resulted in a surge of malware. Conventional detection technologies face difficulties in dealing with constantly changing threats. This thesis investigates the utilisation of artificial intelligence and dimensionality reduction techniques to improve the accuracy of malware detection. Artificial Intelligence, particularly machine learning and deep learning, is capable of identifying patterns even in novel forms of malware. However, the presence of a large number of variables often leads to overfitting and significant computing expenses. This work enhances model efficiency by implementing dimensionality reduction techniques such as PCA, LDA, and t-SNE, which effectively compress the feature space while preserving essential information. The project gathers both harmless and harmful Android applications, extracts their characteristics, using dimensionality reduction techniques, and trains artificial intelligence models such as Support Vector Machines (SVM), Random Forest (RF), and Convolutional Neural Networks (CNN). The findings indicate that the integration of artificial intelligence with dimensionality reduction enhances both the precision and efficiency of the models, rendering them appropriate for use on mobile devices in real-time scenarios. This method improves cybersecurity by providing more flexible and efficient security solutions to safeguard mobile environments. The results highlight the capacity of these technologies to offer strong malware detection systems.

Copyright © 2024 AAKANSHA. This is an open access article distributed under the Creative Commons Attribution License.