In the recent years, cloud databases have gained popularity over the internet transactions. However, Cloud security issues provide new challenges to the internet and cloud users. In the literature, many security techniques have been proposed to enhance the cloud security. In spite of these algorithms, the current cloud systems face a lot of threats on the security of the cloud data. Hence, in this paper we propose a new intelligent intrusion detection system by combining a new optimal feature selection algorithm and a new hybrid multiclass classification algorithm for providing enhanced security to cloud databases. This Hybrid Multiclass Classification algorithm is proposed by combining the intelligent agent based enhanced multiclass support vector machines with Naive Bayes Classifier. The new hybrid feature selection method called Hybrid Genetic based Feature Selection Algorithm (HGFSA) is proposed in this paper for effective classification. Experiments have been carried out on cloud datasets for testing the proposed Genetic based attributes selection algorithm and Hybrid Multiclass Classifier. © Research India Publications.