A Deep Learning Framework for Real-Time Intelligent Decision-Making Systems

Real-time intelligent decision-making remains one of the most demanding challenges in modern artificial intelligence, particularly indynamic, resource-constrained environments where latency and accuracy must be simultaneously optimised. Existing approaches oftensacrifice interpretability or temporal responsiveness for predictive performance, leaving a critical gap in operational deployment. This studyproposes a novel hybrid deep learning framework, termed Temporal Attention-Guided Decision Network (TAG-DN), which integratesconvolutional neural networks (CNN), long short-term memory (LSTM) units, and a multi-head self-attention mechanism to enable precise,low-latency decision-making across sequential data streams. The framework was evaluated on three benchmark datasets—KITTIautonomous driving sequences, UCF-Crime surveillance streams, and a custom industrial IoT sensor dataset comprising 120,000timestamped records. TAG-DN achieved a mean decision accuracy of 94.7%, outperforming state-of-the-art baselines including standardLSTM (88.3%), Transformer-only models (91.2%), and CNN-BiLSTM hybrids (92.6%). Average inference latency was measured at 11.4milliseconds per decision cycle on an NVIDIA RTX 3080 GPU. The attention mechanism improved temporal feature localisation by 18.3%over non-attention baselines. Ablation studies confirmed the complementary contribution of each architectural component. The proposedframework demonstrates strong generalisation across diverse real-time application domains and offers a scalable, deployable solution forintelligent systems requiring rapid, data-driven decisions.Keywords: