This report examines the integration of advanced cryptography and machine learning to improve data privacy and security in cloud computing. It uses AES encryption for data confidentiality and SHA-256 hashing for integrity verification. Machine learning models, including K-Nearest Neighbors (KNN), Isolation Forest, One-Class SVM, Gaussian Mixture Model (GMM), and Autoencoder, were employed for real-time anomaly detection to block anomalous data transfers.

The BETH dataset, reflecting cloud traffic behavior, was used for training and testing. Google Colab’s GPU acceleration helped reduce model training times. KNN achieved the best performance, with 94.57% accuracy and 99.96% precision, making it the most effective for anomaly detection.
Despite limitations like the use of basic encryption methods and the computational constraints of Google Colab, the project provides a solid framework for combining cryptography with machine learning for cloud security. Future work will focus on implementing advanced encryption techniques and scaling the system to larger datasets using platforms like AWS or GCP. This study offers a scalable approach to secure cloud computing, enhancing protection against data breaches.