Questions in causality, control, and reinforcement learning go beyond the classical machine learning task of prediction under i. i. d.
Learning optimal control policies directly on physical systems is challenging since even a single failure can lead to costly hardware damage.
When learning policies for robotic systems from data, safety is a major concern, as violation of safety constraints may cause hardware damage.
In particular, we discuss the family of constraints that enforce safety in the context of a nominal control policy, and expose that these constraints do not need to be accurate everywhere.
Probabilistic models such as Gaussian processes (GPs) are powerful tools to learn unknown dynamical systems from data for subsequent use in control design.
The combination of machine learning with control offers many opportunities, in particular for robust control.
However, these estimates are of a Bayesian nature, whereas for some important applications, like learning-based control with safety guarantees, frequentist uncertainty bounds are required.
An important class of cyber-physical systems relies on multiple agents that jointly perform a task by coordinating their actions over a wireless network.
On the other hand, classical numerical integrators are specifically designed to preserve these crucial properties through time.
Accurate models of mechanical system dynamics are often critical for model-based control and reinforcement learning.
We consider failing behaviors as those that violate a constraint and address the problem of learning with crash constraints, where no data is obtained upon constraint violation.
We present a framework for model-free learning of event-triggered control strategies.
We present a method for automatically identifying the causal structure of a dynamical control system.
When learning to ride a bike, a child falls down a number of times before achieving the first success.
Evaluating whether data streams were generated by the same distribution is at the heart of many machine learning problems, e. g. to detect changes.
Despite the availability of ever more data enabled through modern sensor and computer technology, it still remains an open problem to learn dynamical systems in a sample-efficient way.
In reinforcement learning (RL), an autonomous agent learns to perform complex tasks by maximizing an exogenous reward signal while interacting with its environment.
While safety can only be guaranteed after learning the safety measure, we show that failures can already be greatly reduced by using the estimated measure during learning.
Learning robot controllers by minimizing a black-box objective cost using Bayesian optimization (BO) can be time-consuming and challenging.
Policy gradient methods are powerful reinforcement learning algorithms and have been demonstrated to solve many complex tasks.
Bayesian optimization is proposed for automatic learning of optimal controller parameters from experimental data.
Event-triggered control (ETC) methods can achieve high-performance control with a significantly lower number of samples compared to usual, time-triggered methods.
Soft microrobots based on photoresponsive materials and controlled by light fields can generate a variety of different gaits.
A supervised learning framework is proposed to approximate a model predictive controller (MPC) with reduced computational complexity and guarantees on stability and constraint satisfaction.
Apart from its application for encoding a sequence of observations, we propose to use the compression achieved by this encoding as a criterion for model selection.
Common event-triggered state estimation (ETSE) algorithms save communication in networked control systems by predicting agents' behavior, and transmitting updates only when the predictions deviate significantly.
State-space models (SSMs) are a highly expressive model class for learning patterns in time series data and for system identification.
In practice, the parameters of control policies are often tuned manually.
With this framework, an initial set of controller gains is automatically improved according to a pre-defined performance objective evaluated from experimental data.
To address this issue, we show how a recently published robustification method for Gaussian filters can be applied to the problem at hand.
The contribution of this paper is to show that any Gaussian filter can be made compatible with fat-tailed sensor models by applying one simple change: Instead of filtering with the physical measurement, we propose to filter with a pseudo measurement obtained by applying a feature function to the physical measurement.