However, a single instruction may be insufficient to fully communicate our intent or, even if it is, may be insufficient for an autonomous agent to actually understand how to perform the desired task.
•
We present a method for fast training of vision based control policies on real robots.
•
In the NeurIPS 2018 Artificial Intelligence for Prosthetics challenge, participants were tasked with building a controller for a musculoskeletal model with a goal of matching a given time-varying velocity vector.
•
Standard sequential generation methods assume a pre-specified generation order, such as text generation methods which generate words from left to right.
•
Missing value imputation is a fundamental problem in modeling spatiotemporal sequences, from motion tracking to the dynamics of physical systems.
•
Imitation learning (IL) aims to learn an optimal policy from demonstrations.
•
The current state-of-the-art Scrabble agents are not learning-based but depend on truncated Monte Carlo simulations and the quality of such agents is contingent upon the time available for running the simulations.
We describe MELEE, a meta-learning algorithm for learning a good exploration policy in the interactive contextual bandit setting.
•
Although deep reinforcement learning has achieved great success recently, there are still challenges in Real Time Strategy (RTS) games.
•
People solve the difficult problem of understanding the salient features of both observations of others and the relationship to their own state when learning to imitate specific tasks.