End-to-end learning

From Computer Science Wiki

Exclamation.png This is student work which has not yet been approved as correct by the instructor WIP

Case study notes[1]

Introduction[edit]

End-to-end learning process is a type of Deep_learning process in which all of the parameters are trained jointly, rather than step by step. [2] Furthermore, just like in the case of Deep_learning process, in end-to-end learning process the machine uses previously gained human input, in order to execute its task.[3] This process is specifically prevalent in the autonomous cars industry(our 2018's case study), as this process's benefits fit perfectly with the car's Convolutional neural networks (CNNs).

How does it work or a deeper look[edit]

End-to-end learning process can be separated into two major components (similarly to Deep_learning process). Training component[4] is the first phase, in which the machine records all of the parameters executed by the human operator (through Convolutional neural networks (CNNs)). Inference component[5] then is possible, with the machine acting upon previously gained experience from the Training component of the end-to-end learning process.

The only difference between end-to-end learning process and Deep_learning process is that the end-to-end learning process must collect all of the parameters jointly(at the same time), while Deep_learning process can collect the parameters ether jointly or step by step. Therefore, every end-to-end learning process is Deep_learning process , but not every Deep_learning process is end-to-end learning process.

Examples[edit]

End-to-end learning process is specifically prevalent in the autonomous cars industry(our 2018's case study), as this process benefits fit perfectly with the car's Convolutional neural networks (CNNs). As the autonomous car jointly receives multiple parameters through Convolutional neural networks (CNNs), it is beneficial to use end-to-end learning process which is able to Train or Infer upon the parameters.

One example could be: "the autonomous car is required to turn right to the civilized area from a highway", as there is a certain speed limit the car needs to adjust its speed accordingly, while at the same time the car needs to turn to the right as well. In this situation end-to-end learning lets the car execute the correct Inference based upon multiple receive parameters.


Pictures, diagrams[edit]

Training-624x291.png

As one can see the masterfully edited picture in paint by the true paint prodigy on the right. The circled parameters are assessed jointly(at the same time), while the entire thing still remains to be Deep_learning process. As the received parameters are assessed jointly within this Deep_learning process, this process can be classified as a end-to-end learning process as well.

External links[edit]

  • It would be helpful
  • to include many links
  • to other internet resources
  • to help fellow students
  • Please make sure the content is good
  • and don't link to a google search results, please

References[edit]