Title | ||
---|---|---|
Robot Learning by Collaborative Network Training: A Self-Supervised Method using Ranking |
Abstract | ||
---|---|---|
We introduce Collaborative Network Training - a self-supervised method for training neural networks with aims of: 1) enabling task objective functions that are not directly differentiable w.r.t. the network output; 2) generating continuous-space actions; 3) more direct optimization for achieving a desired task; 4) learning parameters when a process for measuring performance is available, but labeled data is unavailable. The procedure involves three randomly initialized independent networks that use ranking to train one another on a single task. The method incorporates qualities from ensemble and reinforcement learning as well as gradient free optimization methods such as Nelder-Mead. We evaluate the method against various baselines using a variety of robotics-related tasks including inverse kinematics, controls, and planning in both simulated and real-world environments. |
Year | DOI | Venue |
---|---|---|
2019 | 10.5555/3306127.3331839 | adaptive agents and multi-agents systems |
Keywords | Field | DocType |
Robot learning,collaborative network training,controls | Robot learning,Ranking,Inverse kinematics,Computer science,Differentiable function,Artificial intelligence,Labeled data,Collaborative network,Artificial neural network,Machine learning,Reinforcement learning | Conference |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Mason Bretan | 1 | 18 | 2.28 |
Sageev Oore | 2 | 101 | 18.63 |
Sanan Siddharth | 3 | 16 | 2.27 |
Larry P. Heck | 4 | 1096 | 100.58 |