On the Design of Autonomous Agents From Multiple Data Sources

Emiland, Garrabe,; Russo, Giovanni

doi:10.1109/lcsys.2021.3086058

Cited by 10 publications

(19 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this context, a key challenge is to design agents able to make decisions by re-using this distributed information, without having to necessarily gather new data. A way to tackle this challenge might be the design of open and transparent crowdsourcing mechanisms for autonomous decision-makers [115,116]. These mechanisms, besides giving performance guarantees should also be able to handle cooperative/competitive behaviors among peers [171].…”

Section: Concluding Discussionmentioning

confidence: 99%

“…For example, in e.g. [115,116] for a shared economy/smart cities application, the KL-divergence component models the fact that an autonomous car would like to stay on a route that accommodates the preferences of the passengers and the additional cost instead depends on road (and/or parking) conditions.…”

Section: Fully Probabilistic Designmentioning

confidence: 99%

“…The optimal solution to the KLC problem (29) and to the FPD problem (25) have a similar structure and can be both interpreted in terms of twisted kernels. See [116] for an explicit link between the two solutions.…”

Section: Kl-controlmentioning

confidence: 99%

See 2 more Smart Citations

Probabilistic design of optimal sequential decision-making algorithms in learning and control

Emiland¹,

Russo²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

This survey is focused on certain sequential decision-making problems that involve optimizing over probability functions. We discuss the relevance of these problems for learning and control. The survey is organized around a framework that combines a problem formulation and a set of resolution methods. The formulation consists of an infinite-dimensional optimization problem. The methods come from approaches to search optimal solutions in the space of probability functions. Through the lenses of this overarching framework we revisit popular learning and control algorithms, showing that these naturally arise from suitable variations on the formulation mixed with different resolution methods. A running example, for which we make the code available, complements the survey. Finally, a number of challenges arising from the survey are also outlined.

show abstract

Section: Concluding Discussionmentioning

confidence: 99%

Section: Fully Probabilistic Designmentioning

confidence: 99%

See 1 more Smart Citation

Probabilistic design of optimal sequential decision-making algorithms in learning and control

Emiland¹,

Russo²

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Related works: we briefly survey a number of works related to the results and methodological framework of this paper. The design of context-aware switches between multiple datasets for autonomous agents has been recently considered in [3], [4], where the design problem, formalized as a data-driven control (DDC) problem, did not take into account safety requirements. Results in DDC include [6]- [8], which take a behavioral systems [9] perspective, [10], which finds data-driven formulas towards a model-free theory of geometric control.…”

Section: Introductionmentioning

confidence: 99%

Optimal Decision-Making for Autonomous Agents via Data Composition

Emiland¹,

Lamberti²,

Russo³

2023

Preprint

View full text Add to dashboard Cite

We consider the problem of designing agents able to compute optimal decisions by composing data from multiple sources to tackle tasks involving: (i) tracking a desired behavior while minimizing an agent-specific cost; (ii) satisfying safety constraints. After formulating the control problem, we show that this is convex under a suitable assumption and find the optimal solution. The effectiveness of the results, which are turned in an algorithm, is illustrated on a connected cars application via in-silico and in-vivo experiments with real vehicles and drivers. All the experiments confirm our theoretical predictions and the deployment of the algorithm on a real vehicle shows its suitability for in-car operation.

show abstract

“…An online version of these algorithms has been proposed in Guan et al (2014): in such a work, by leveraging an average cost formulation, the probability mass function for the state transitions is found. Finally, we also recall here Russo (2021); Garrabe and Russo (2022), where policies are obtained from the minimization of similar costs by leveraging multiple, specialized, datasets. These last papers, together with Gagliardi and Russo (2021), introduce constraints to the probabilistic formulation.…”

Section: Introductionmentioning

confidence: 99%

Discrete fully probabilistic design: towards a control pipeline for the synthesis of policies from examples

Ferrentino¹,

Chiacchio²,

Russo³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

We present a discretized design that expounds an algorithm recently introduced in Gagliardi and Russo (2021) to synthesize control policies from examples for constrained, possibly stochastic and nonlinear, systems. The constraints do not need to be fulfilled in the possibly noisy example data, which in turn might be collected from a system that is different from the one under control. For this discretized design, we discuss a number of properties and give a design pipeline. The design, which we term as discrete fully probabilistic design, is benchmarked numerically on an example that involves controlling an inverted pendulum with actuation constraints starting from data collected from a physically different pendulum that does not satisfy the system-specific actuation constraints.

show abstract

On the Design of Autonomous Agents From Multiple Data Sources

Cited by 10 publications

References 18 publications

Probabilistic design of optimal sequential decision-making algorithms in learning and control

Probabilistic design of optimal sequential decision-making algorithms in learning and control

Optimal Decision-Making for Autonomous Agents via Data Composition

Discrete fully probabilistic design: towards a control pipeline for the synthesis of policies from examples

Contact Info

Product

Resources

About