We consider the problem of learning to follow a desired trajectory when given a small number of demonstrations from a sub-optimal expert. We present an algorithm that (i) extracts the—initially unknown—desired trajectory from the sub-optimal expert’s demonstrations and (ii) learns a local model suitable for control along the learned trajectory. We apply our algorithm to the problem of autonomous helicopter flight. In all cases, the autonomous helicopter’s performance exceeds that of our expert helicopter pilot’s demonstrations. Even stronger, our results significantly extend the state-of-the-art in autonomous helicopter aerobatics. In particular, our results include the first autonomous tic-tocs, loops and hurricane, vastly superior performance on previously performed aerobatic maneuvers (such as in-place flips and rolls), and a complete airshow, which requires autonomous transitions between these and various other maneuvers. 1.

We present PROCAB, an efficient method for Probabilistically Reasoning from Observed Context-Aware Behavior. It models the context-dependent utilities and underlying reasons that people take different actions. The model generalizes to unseen situations and scales to incorporate rich contextual information. We train our model using the route preferences of 25 taxi drivers demonstrated in over 100,000 miles of collected data, and demonstrate the performance of our model by inferring: (1) decision at next intersection, (2) route to known destination, and (3) destination given partially traveled route.

Partially Observable Markov Decision Processes (POMDPs) have succeeded in planning domains that require balancing actions that increase an agent’s knowledge and actions that increase an agent’s reward. Unfortunately, most POMDPs are defined with a large number of parameters which are difficult to specify only from domain knowledge. In this paper, we present an approximation approach that allows us to treat the POMDP model parameters as additional hidden state in a “model-uncertainty ” POMDP. Coupled with model-directed queries, our planner actively learns good policies. We demonstrate our approach on several POMDP problems. 1.

We present a Bayesian framework for explaining how people reason about and predict the actions of an intentional agent, based on observing its behavior. Action-understanding is cast as a problem of inverting a probabilistic generative model, which assumes that agents tend to act rationally in order to achieve their goals given the constraints of their environment. Working in a simple sprite-world domain, we show how this model can be used to infer the goal of an agent and predict how the agent will act in novel situations or when environmental constraints change. The model provides a qualitative account of several kinds of inferences that preverbal infants have been shown to perform, and also fits quantitative predictions that adult observers make in a new experiment. 1

Learning by imitation represents an important mechanism for rapid acquisition of new behaviors in humans and robots. A critical requirement for learning by imitation is the ability to handle uncertainty arising from the observation process as well as the imitator’s own dynamics and interactions with the environment. In this paper, we present a new probabilistic method for inferring imitative actions that takes into account both the observations of the teacher as well as the imitator’s dynamics. Our key contribution is a nonparametric learning method which generalizes to systems with very different dynamics. Rather than relying on a known forward model of the dynamics, our approach learns a nonparametric forward model via exploration. Leveraging advances in approximate inference in graphical models, we show how the learned forward model can be directly used to plan an imitating sequence. We provide experimental results for two systems: a biomechanical model of the human arm and a 25-degrees-of-freedom humanoid robot. We demonstrate that the proposed method can be used to learn appropriate motor inputs to the model arm which imitates the desired movements. A second set of results demonstrates dynamically stable full-body imitation of a human teacher by the humanoid robot. 1

Many situations arise in which an interested party’s utility is dependent on the actions of an agent; e.g., a teacher is interested in a student learning effectively and a firm is interested in a consumer’s behavior. We consider an environment in which the interested party can provide incentives to affect the agent’s actions but cannot otherwise enforce actions. In value-based policy teaching, we situate this within the framework of sequential decision tasks modeled by Markov Decision Processes, and seek to associate limited rewards with states that induce the agent to follow a policy that maximizes the total expected value of the interested party. We show value-based policy teaching is NP-hard and provide a mixed integer program formulation. Focusing in particular on environments in which the agent’s reward is unknown to the interested party, we provide a method for active indirect elicitation wherein the agent’s reward function is inferred from observations about its response to incentives. Experimental results suggest that we can generally find the optimal incentive provision in a small number of elicitation rounds.

Decision-theoretic models have become increasingly popular as a basis for solving agent and multiagent problems, due to their ability to quantify the complex uncertainty and preferences that pervade most nontrivial domains. However, this quantitative nature also complicates the problem of constructing models that accurately represent an existing agent or multiagent system, leading to the common question, "Where do the numbers come from?" In this work, we present a method for exploiting knowledge about the qualitative structure of a problem domain to automatically derive the correct quantitative values that would generate an observed pattern of agent behavior. In particular, we propose the use of piecewise linear functions to represent probability distributions and utility functions with a structure that we can then exploit to more efficiently compute value functions. More importantly, we have designed algorithms that can (for example) take a sequence of actions and automatically generate a reward function that would generate that behavior within our agent model. This algorithm allows us to efficiently fit an agent or multiagent model to observed behavior. We illustrate the application of this framework with examples in multiagent modeling and social simulation, using decision-theoretic models drawn from the alphabet soup of existing research (e.g., MDPs, POMDPs, Dec-POMDPs, Com-MTDPs).

In multiagent environments, forms of social learning such as teaching and imitation have been shown to aid the transfer of knowledge from experts to learners in reinforcement learning (RL). We recast the problem of imitation in a Bayesian framework.

doi:10.1145/1538788.1538812 Autonomous helicopter flight is widely regarded to be a highly challenging control problem. As helicopters are highly unstable and exhibit complicated dynamical behavior, it is particularly difficult to design controllers that achieve high performance over a broad flight regime. While these aircraft are notoriously difficult to control, there are expert human pilots who are nonetheless capable of demonstrating a wide variety of maneuvers, including aerobatic maneuvers at the edge of the helicopter’s performance envelope. In this paper, we present algorithms for modeling and control that leverage these demonstrations to build high-performance control systems for autonomous helicopters. More specifically, we detail our experiences with the Stanford Autonomous Helicopter, which is now capable of extreme aerobatic flight meeting or exceeding the performance of our own expert pilot. 1.

Abstract not found