Improving Cloud Data Security with Alexnet Learning Model Compare Stochastic Gradient Descent with Accuracy

The suggested method of protecting cloud data involves setting up a server to function as a third party, similar to Alexnet, and using Advanced Encryption Standard for data stored in private clouds. Stochastic Gradient Descent is used to calculate accuracy the user and the cloud in order to ensure security between the two parties. Supplies and Procedures: In this work, two groups—AlexNet and Stochastic Gradient Descent—are compared. For statistical analysis, 1300 instruction and 423 testing dataset samples from the UNSW-NB15 dataset were used. N=10, an alpha value of 0.05, a G-Power value of 0.8%, and a 95% confidence interval were employed in the ClinCalc software for each approach. Findings and Discussion: Novel Alexnet (91.081%) exhibits better accuracy than SGD (87.9396%). The significance value, as determined by looking at the Independent Sample t-Test findings, is 0.011(p<0.05). It is statistically significant as a result. Conclusion: Based on the research findings, the Novel AlexNet classifier method outperforms the SGD in terms of improving the precision of cloud data that contains sensitive information using the UNSW-NB15 dataset.