Overview of Enabling End to End Machine Learning
An algorithm can solve the whole function on its own. There is no need for more human intervention, such as model swapping or new data marking. For example, end-to-end driving means that the neural network is only analyzing objects to find out how to change the steering order.