This research study analyzes the performance of machine learning-based algorithms in terms of efficiency in classifying security attacks in cloud computing systems. The proposed research consists of two phases: identifying and wrangling data as per the required format and performing analysis of different machine learning algorithms based on security attack detection and classification using the CSIC HTTP dataset. It contains a set of static and dynamic attacks created using connected objects. Security attacks are detected and analyzed on machine learning algorithms, like k-nearest Neighbor, random forest, and support vector machine. This research helps to find out the set of illegal requests that deteriorate the privacy of shared data. The performance of algorithms is explored through a confusion matrix. During analysis KNN has shown good balance between precision and recall, which is significant in the classification task over Random Forest and support Vector Machine models for different-sized test data.