A PhD Research Project: Hierarchical Decision-Making Under Uncertainty
Recently Published Research: Reinforcement Learning Under Epistemic Uncertainty
Maksim Loginov
Open to ML / Data Engineering roles, research collaborations, and industrial AI prototype work.
I build simulation-based prototypes for maintenance optimization, operator-in-the-loop decision loops, and uncertainty-aware diagnostics — combining hierarchical MARL research with practical Python / SQL engineering.
I am building a research-to-prototype path for industrial decision-support systems: from formal models of uncertainty and reinforcement learning to small, testable simulation environments. The current focus is maintenance optimization, operator-in-the-loop decision workflows, and diagnostic models that remain useful when data is incomplete or conflicting.
Published research support for CCM diagnostics under uncertainty — with journal PDF and live demo.
Belief Functions · VIVT Journal · Demo available
PDF · Demo · Case study
Simulation-based decision-support prototype: hierarchical MARL, HITL feedback, constrained policies.
MARL · HRL · CMDP · HITL · Maintenance Simulation
My research centers on decision-making under uncertainty: hierarchical and multi-agent reinforcement learning for industrial maintenance, human-in-the-loop approval loops, and evidential (Dempster–Shafer) models for diagnostics. I aim to connect formal decision models with simulation environments and small engineering prototypes.
Recently Published Research: Reinforcement Learning Under Epistemic Uncertainty
Reinforcement Learning for Decision-Making in Complex Systems
Phd article on hierarchical multi-agent reinforcement learning framework for maintenance and repair (MRO) optimizatio...
ML Engineering, Data Engineering, Research Engineering, and industrial AI prototype roles.
Collaboration on RL, MARL, HITL, Dempster–Shafer theory, and decision-making under uncertainty.
Simulation-based studies, maintenance optimization, diagnostics, and operator-in-the-loop decision support.