Hello,
I’d like to train a LSTM network in order to remember the steps that compose a trajectory to reach an object. So far, I’ve been using RL, so I’m thinking in terms of states and action . For instance, is it possible that the agent would get an initial state, produce an action and recognize the sequence and proceed with it ? Or would it be better to train it to produce only states ? But then, how could I control it ?
Thanks !
Do you need an LSTM for that ? Why not just saving the positions somewhere and backtracking them ?
Hi, thanks for answering,
What do you mean ? Having a simple ff network output next expected position given state?
You want your robot to remember a trajectory to reach an object ? So why do you need a Neural network for that ?
Well, because I want to see whether it is possible to be able to generalize, hence avoiding to compute pseudo-inverse and all the other calculations necessary for planning.
Are you looking at discrete data for your positions ? and then interpolate the path ? or a continuous path ?
I would look at the possibility to learn states personally