Abstract

An internet-wide computer network is made up of several machines that are connected, whether the network is built using traditional infrastructure or cloud computing technology. All of these goods, services and programs come together in one technical framework, together with a large amount of database storage, creating a valuable resource variety. Internet services that enable consumers to make use of the vast possibilities provided by cloud technology facilitate access to these products. Cloud computing's many benefits include low cost, easy-to-use self-service accessibility, built-in flexibility, smooth scalability, and efficient use of virtualised resources, which are driving creativity and effectiveness in contemporary technology paradigms. However, there are several security vulnerabilities in the cloud computing (CC) setup. Trojan horses (TH) are among the most frequent security issues in the context of computing in the cloud. The TH can compromise services offered by cloud computing and harm virtual machines, apps and resources in the cloud infrastructure. TH assaults could be hazardous, intricate and hard to identify. We begin by assembling an extensive dataset with TH Infections for Securing Financial Cloud data. We proposed the Adaptive Butterfly-optimised Hierarchical Random Forest (ABO-HRF) technique for accurately estimating TH detection. The suggested approach improves the accuracy (89.56%), precision (88.32%), recall (82.54%), and F1 score (84.68%) compared with the current methods.