Quantum dot (QD) superlattices are promising materials for optoelectronic devices, but optimizing their photonic properties remains a complex challenge. We developed a machine learning (ML)-driven optimization framework to predict and optimize key photonic properties of QD superlattices. Our approach combines quantum mechanical models with ML algorithms to forecast the behavior of QD structures based on their physical parameters. We trained a neural network model on a dataset of 1000 simulated QD configurations, achieving a mean absolute error (MAE) of 0.05 eV for photonic bandgap frequency and 10 nm for emission wavelength. Optimization results showed significant improvements in optical efficiency (up to 25%) and photonic bandgap (up to 15%) across a range of QD configurations. Sensitivity analysis revealed that lattice constant and inter-dot spacing are the primary drivers of variability in the photonic bandgap. Our findings demonstrate the potential of ML-driven optimization for designing high-performance QD-based devices, with implications for optoelectronics, photonics, and energy conversion systems. This study provides a scalable methodology for optimizing nanomaterials, enabling the rapid design and deployment of next-generation optoelectronic devices.