DDoS attacks, a type of network security threat, have spread to other technologies like cloud computing, IoT, and edge computing. According to the behavior, DDoS attacks use all available resources, including memory, CPU, and network, to bring down the victim’s machine or server. Although several protective mechanisms have been offered, they are ineffective as attackers may easily teach themselves using automated tools. This paper presents, DDoS attack detection in cloud computing using geographically weighted artificial neural network (DDos-ADCC-GWANN) to predict and prevent the DDos attack in cloud computing. Initially the data are transmitted as input data after being extracted from the NSL-KDD data collection. The supplied data is then passed into pre-processing. During pre-processing, to remove noise, Unsharp Mask Guided Filtering (UMGF) is employed. The outcome of pre-processingis transmitted to extract features using the Two-sided Offset Quaternion Linear Canonical Transform (TOQLCT) to extract a flow time, flags, protocol name, and service name characteristics that are employed to categorize samples as normal or attack. (TOQLCT). After that, extracted features are given to geographically weighted artificial neural network (GWANN) for classifying the input data as normal or DDos attack and TOQLCT is optimized with Binary Waterwheel Plant Optimization algorithm (BWPO) to get better results in prediction. The proposed DDos-ADCC-GWANN approach is implemented in MALTAB platform. When compared to other methods such as the Optimized Extreme Learning Machine for Identifying DDoS Attacks in Cloud Computing (OEKL-DDos-CC), the performance of the suggested DDos-ADCC-GWANN methodology obtains 25.45, 19.12, and 27.11% high accuracy, and 15.36, 21.55, and 18.74% greater specificity. In the direction of employing deep learning for DDoS attack detection (DDos-AD-DL) and a DDoS attack classification and prediction method according to machine learning (ML-CP-DDos) methods, respectively.