Real-time big data analytics is crucial for industries that handle high-velocity data streams, such as IoT, finance, e-commerce, and social media. Traditional big data processing frameworks often struggle to adapt to the dynamic nature of real-time data, leading to inefficiencies, increased latency, and higher operational costs. This research proposes an Adaptive Real-Time Big Data Processing Framework that integrates a Random Forest model for load prediction with Q-learning for dynamic resource allocation. The framework adjusts in real-time based on predictive insights and learned policies, optimizing system performance while minimizing costs.
Experimental results demonstrate that the framework reduces latency by 15%, improves throughput by 20%, compared to traditional methods. The Random Forest model achieved an efficiency of 93.62%, while the Q-learning approach further achieved 94.86%, reflecting superior resource utilization. These improvements indicate the framework's effectiveness in dynamically adapting to real-time data, ensuring high performance and cost-efficiency.