@article{YUJOR, author = {Santosh Maurya and Suraj Malik and Neeraj Kumar and Hare Ram Singh}, title = { A Machine Learning Based CIDS Model for Intrusion Detection to Ensure Security Within Cloud Network}, journal = {Yugoslav Journal of Operations Research}, year = {2025}, keywords = {}, abstract = {Cloud computing has changed quickly in recent years, and security issues have become more prevalent. Through the Internet, cloud computing has the amazing ability to deliver powerful, versatile, adaptable, affordable, and manageable assets that are always on the go. The potential of hardware resources is maximized by cloud computing through efficient and shared use. Data and service availability are challenges raised by a number of cloud computing difficulties. To improve cloud security for consumers and service providers, a multitude of security services are required. In order to provide security within the cloud and prevent load imbalance scenarios, this study suggests using Cloud Intrusion Detection Systems (CIDS). Cloud security vulnerabilities include denial of service, scanning, malware code injection, viruses, worms, and password cracking. If these attacks are not discovered in time, they might harm the company's finances and image. Our idea through this work aims to protect the cloud from these types of attacks and to accurately detect and anticipate them early on. Previous research projects have noted that when dimensionality reduction is used in conjunction with data mining (DM), machine learning (ML) techniques have been proven to perform better. The authors proposed a CIDS by choosing relevant characteristics using relevant feature reduction techniques, then feeding this subset of data through the ML tool in order to develop such a robust model to assure a cloud network. Python and the Scikit-Learn program are used to simulate the proposed model. Using the KDDcup99 dataset as a benchmark, the simulation experiment's results were assessed using a variety of performance assessment criteria, including precision, recall, F-Score, detection ratio, RoC curve, etc. Our recommended methodology produced simulation results that were more efficient and on par with a number of other approaches. It has been noted that the ML-based proposed model was found to be sufficiently capable of safeguarding cloud-based data by identifying potentially suspicious user behavior, protecting the cloud network from threats, and demonstrating superior performance in true prediction and early intrusion detection, which led to a reduction in computational costs.}, issn = {2334-6043}, doi = {10.2298/YJOR240315036K}, url = {https://yujor.fon.bg.ac.rs/index.php/yujor/article/view/1327} }