V.Satyanarayana, Dr P.Chiranjeevi 03 06 Cloud computing has become the backbone of modern digital infrastructure enabling scalable storage and seamless data transmission across distributed networks However the increasing volume of sensitive data transmitted over cloud platforms has made them prime targets for cyberattacks such as data interception Distributed Denial of Service DDoS and advanced persistent threats APTs Traditional security mechanisms including static encryption protocols and signaturebased intrusion detection systems are insufficient in addressing dynamic and evolving threats This paper proposes an advanced machine learningbased framework for enhancing data transmission security in cloud environments The system integrates supervised and unsupervised learning models for anomaly detection realtime traffic analysis and adaptive encryption mechanisms A hybrid approach combining Random Forest Support Vector Machine and Neural Networks is utilized to improve detection accuracy and reduce false positives The system is evaluated using benchmark datasets such as NSLKDD achieving an accuracy of up to 98 The proposed model demonstrates improved resilience scalability and adaptability compared to conventional approaches Copyright (c) Journal of Science Engineering Technology and Management Science. All rights reserved 2026