The research involves exploring various security issues in the cloud environment at with respect to three aspects confidentiality, integrity and availability and analyzes their impacts. The research proposes secure authentication in the cloud computing using image sequencing passwords. Then a secure data classification model using novel boosting supervised machine learning approach. In this, data is classified according to its sensitivity level. Then encrypting only, the data which is required to be secure using a hybrid privacy preserving based image steganography technique in order to reduce the data hiding time and ensure the integrity and confidentiality of data using Hashing approach.