Hybrid LiFi/WiFi networks leverage the high-speed potential of optical wireless links alongside the coverage reliability of RF technologies, but LiFi’s limited range and blockage sensitivity often trigger frequent handovers. Accurate prediction of Signal-To-Interference-plus-Noise Ratio (SINR) and Signal-To-Noise Ratio (SNR) from user position data can improve access point (AP) selection and reduce unnecessary handovers. This paper compares Supervised Learning (SL), Reinforcement Learning (RL), and a hybrid SL+RL approach using SINR/SNR measurements from a LiFi/WiFi environment, evaluating prediction accuracy, throughput, handover rate, and fairness.