Risk-Aware GUMDPs Optimize Decisions with Entropic Risk Measures

Pedro P. Santos, F\'abio Vital, Alberto Sardinha, Francisco S. Melo· July 13, 2026 View original

Summary

This paper introduces Risk-Aware General-Utility Markov Decision Processes (GUMDPs) that optimize a risk measure of objective values, focusing on the entropic risk measure (ERM). It proposes an MCTS-based approach to solve these GUMDPs, demonstrating its success across various tasks.

Traditional Markov Decision Processes (MDPs) often focus on optimizing expected outcomes, which may not adequately capture risk preferences in decision-making. This research introduces Risk-Aware General-Utility Markov Decision Processes (GUMDPs), a framework where agents optimize a risk measure of the distribution of objective values, rather than just the mean. The objective function itself depends on the frequency of state visitation induced by the agent's policy. The paper specifically focuses on the entropic risk measure (ERM), which allows decision-makers to explicitly trade off expected performance with risk aversion. This enables a richer set of objectives to be considered within the GUMDP framework. To solve these complex risk-aware GUMDPs, the researchers propose an online planning approach based on Monte Carlo Tree Search (MCTS). This method is provably capable of solving risk-aware GUMDPs to any desired accuracy. Experimental results across diverse tasks, including standard MDPs, maximum state entropy exploration, imitation learning, and multi-objective MDPs, demonstrate the success of this approach in optimizing for a spectrum of risk-aware behaviors.

Why it matters

For professionals in domains where decisions involve significant uncertainty and potential high-impact consequences, this framework provides a robust way to incorporate risk aversion directly into AI planning and control systems, leading to more prudent and strategically aligned automated decisions.

How to implement this in your domain

  1. 1Apply Risk-Aware GUMDPs to develop AI agents for financial trading or portfolio management, incorporating specific risk tolerance levels.
  2. 2Utilize the MCTS-based solver for planning in autonomous systems where safety and risk mitigation are paramount, such as robotics or self-driving cars.
  3. 3Integrate entropic risk measures into reinforcement learning models for critical infrastructure management or resource allocation.
  4. 4Explore the framework for multi-objective optimization problems where balancing different risks and rewards is essential.

Who benefits

Financial ServicesInsuranceAutonomous SystemsLogisticsHealthcare

Key takeaways

  • Risk-Aware GUMDPs allow AI agents to optimize for risk measures beyond just expected outcomes.
  • The framework specifically focuses on the entropic risk measure (ERM) for explicit risk aversion.
  • An MCTS-based online planning approach can provably solve these complex risk-aware GUMDPs.
  • This method enables more sophisticated risk-aware behaviors in diverse AI planning tasks.

Original post by Pedro P. Santos, F\'abio Vital, Alberto Sardinha, Francisco S. Melo

"arXiv:2607.09298v1 Announce Type: new Abstract: We study general-utility Markov decision processes (GUMDPs) with risk-aware objectives. In this framework, an agent aims to optimize a risk measure of the distribution of objective values, where the objective function depends on the…"

View on X

Originally posted by Pedro P. Santos, F\'abio Vital, Alberto Sardinha, Francisco S. Melo on X · view source

Want to go deeper?

Turn these trends into skills with Learnijoy's hands-on AI & tech courses.

Explore courses

More in AI Research

AI ResearchAI Engineering & DevTools

Data-Efficient Deep Learning Guidelines for Inertial Sensor Classification

This study provides empirical guidelines for estimating the minimum training set size needed for deep learning models in inertial sensor classification tasks. It reveals that accuracy follows a consistent logarithmic growth pattern, allowing for data-efficient planning of recording campaigns and achieving practical stability with fewer samples than traditionally assumed.

Ofir Kruzel, Itzik KlienJul 13, 2026
AI Engineering & DevToolsAI Research

On-Device Adaptive AI Boosts EV Battery Power Prediction

Researchers developed a novel approach for on-device learning in electric vehicles (EVs) that continuously adapts pretrained battery power prediction models to new data. This method significantly improves forecasting performance, reducing mean absolute errors by up to 14.88% with offline adaptation and 7.49% with online adaptation.

Avik Bhatnagar, Anton Paule, Tobias Schuermann, Sebastian Reiter, Oliver BringmannJul 13, 2026
AI ResearchAI Engineering & DevTools

New Differentiable Logic Networks Outperform Fixed-Connection Models

Researchers introduce a novel method for optimizing connections in deep differentiable logic gate networks (LGNs) and lookup table networks (LUTNs), achieving superior performance with significantly fewer gates. The approach allows for parallel learning of optimal gate types and LUT entries, demonstrating improved accuracy on benchmarks like MNIST.

Wout Mommen, Lars Keuninckx, Matthias Hartmann, Werner Van Leekwijck, Piet WambacqJul 13, 2026