The rapid growth of cloud computing has made efficient resource management essential to ensure performance, scalability, and cost-effectiveness. Traditional approaches to cloud resource scaling, such as manual provisioning and rule-based or reactive autoscaling, often fail to meet the dynamic requirements of modern applications. These methods either over-provision resources, resulting in unnecessary costs, or under-provision them, leading to performance degradation and service-level agreement (SLA) violations. To overcome these limitations, this study proposes a Predictive Model for Cloud Resource Scaling Using Machine Learning Techniques that leverages workload forecasting and intelligent decision-making to achieve proactive, adaptive resource management. The proposed model employs a multi-layered architecture. A data collection layer gathers system metrics such as CPU utilization, memory consumption, network throughput, and application latency. These inputs are processed through feature engineering and time-series analysis, enabling the identification of workload patterns and trends. At the core, a machine learning prediction engine—utilizing algorithms such as Long Short-Term Memory (LSTM), Random Forests, and Reinforcement Learning—forecasts resource demand over short- and long-term horizons. Predictions are fed into a decision engine, which applies optimization strategies to determine scaling actions that balance performance, cost efficiency, and SLA compliance. Finally, an execution layer integrates with cloud orchestration tools (e.g., Kubernetes Horizontal Pod Autoscaler, AWS Auto Scaling) to enforce scaling decisions in real time. Key benefits of the model include improved SLA compliance, reduced operational costs, enhanced system responsiveness, and support for multi-resource scaling across hybrid and multi-cloud environments. However, challenges such as prediction accuracy, cold-start scenarios, and integration complexity remain. By shifting cloud resource management from reactive to predictive, this model offers a pathway toward autonomous, intelligent, and sustainable cloud ecosystems capable of adapting to evolving application and workload demands.