Directions in QML avoiding Barren Plateaus

In recent years, most Quantum Machine Learning (QML) algorithms have been built around variational approaches. However, their limitations—particularly those associated with Barren Plateaus (BPs)—raise concerns about their suitability for QML tasks. This talk explores alternative techniques that could offer better performance in QML by mitigating the impact of BPs. Special attention is given to Bayesian inference methods and their potential applications within the quantum setting.