My goal is to write a program that automatically selects a routing out of multiple proposed options.

The data consists out of the multiple proposed options with each the attributes time, costs and if there is a transhipment and also which of the options was selected.

Example of data:

My idea at the moment is that I have to apply so type of inference to learn which attribute (time, costs, transhipment) has the highest impact on how to choose the best option. But I don't know exactly where to start with this.

Is there a "best" ML algorithm for this? Or how should I approach this?

The dataset currently consists out of 1000 samples in case if this is important.

Thanks in advance for your responses.

  • $\begingroup$ Some clarification needed: 1) Are the choices sequential i.e. choose one option leads to another state where the decision needs to be made again? Or is each choice a one-off, independent one? 2) Is the goal to select the best of 4 choices at any one time? If so, do you have a single consistent measure of "best" e.g. is it the lowest cost, or some combination of time, costs and "transshipment"? Or are you trying to modify something else e.g. you can control time, costs and transshipment and want to make your item the chosen one? $\endgroup$ – Neil Slater Apr 26 at 11:18
  • $\begingroup$ @NeilSlater 1) The choice is independent 2) It probably is a combination of all three, but some attributes might be weighted more than others. And no, I am not trying to modify time, costs and transshipments. Those attributes are given for each option. Please let me know if my question is still unclear. Thank you $\endgroup$ – Nui Apr 26 at 11:50
  • $\begingroup$ Thanks for the clarification. If you are trying to discover the weightings, then you must have some other measure of overall success? If you compare two different results after selecting an option, how do you know which is better? Is your data a list of the best choice out of four for each set that happened in the past, or is it something else? $\endgroup$ – Neil Slater Apr 26 at 12:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.