Machine learning is often based on neural networks and these machine learning methods are frequently referred to black-box that symbolize the transfer functions like inputs and outputs. I would like to learn about this black-box machine learning. Can someone explain to me what is black-box machine learning?