A modern technology which enables users to construct services on demand is cloud computing. Due to its self-service functionality and on demand offerings, cloud computing has seen success. Some security-related risks are brought on by remote cloud-based data evaluation and maintenance, particularly with DDoS Attacks. These assaults are the result of intrusion efforts or hacker attempts to access data stored on the main server or moving between the client and server. Without the data owner's knowledge, the attacker acquires the data and alters it as needed. For the widespread adoption of cloud computing, it is crucial to develop defences against this assault. Over the past few years, intrusion detection (ID) in a cloud setting has drawn a lot of attention. Machine learning-based ID methods are among the most recent techniques that enable us to find unidentified assaults. However, building a cloud based ID system (IDS) that is resistant to a broad variety of unknown attacks remains difficult due to the scarcity of malicious samples and the quick evolution of diverse attacks. This study creates a deep feedforward neural network algorithm-based intruder detection system using Bat optimization. The CICDDOS2019 dataset is used by the suggested technique to analyze performance.