Our Work

More notebooks

Code accompanying the paper "Worst-Case LTI Perturbations of Finite-Horizon LQR: A Performance Estimation Problem Approach" by Anthony Couthures, Gianluca Bianchin and Julien M. Hendrickx (I…

Code accompanying the paper "Worst-Case LTI Perturbations of Finite-Horizon LQR: A Performance Estimation Problem Approach" by Anthony Couthures, Gianluca Bianchin and Julien M. Hendrickx (ICTEAM, UCLouvain). The repository provides a Python/Jupyter implementation for quantifying the finite-horizon robustness of an LQR controller under norm-bounded LTI parametric uncertainties. The worst-case search is reformulated within the Performance Estimation Problem (PEP) framework as a rank-constrained semidefinite program over a Gram matrix. Dropping the rank constraint gives an upper bound. When the relaxed solution has rank one, the bound is exact and the worst-case initial condition and perturbation matrices can be read off directly. Otherwise, a valid LTI worst case is recovered with a Burer–Monteiro rank-1 factorization and a Monte-Carlo-seeded nonlinear program. Contents of this release: Solvers for the nominal LQR, the PEP/Gram-matrix SDP, the Burer–Monteiro QCQP, and Monte-Carlo, NLP and pseudospectral heuristics. An interactive ipywidgets dashboard to vary the system, cost and uncertainty parameters in real time. Scripts that reproduce all figures of the paper: spectral targeting and worst-case trajectories, worst-case pole migration as the horizon grows, and the comparison between IQC (LTV) bounds, relaxed PEP, rank-1 PEP, Monte-Carlo samples and the pseudospectral heuristic. Requirements: Python 3.9+, NumPy, SciPy, CVXPY, Matplotlib, ipywidgets. MOSEK is recommended (free academic license); SCS is used as a fallback. Released under the MIT License.

The failure of traditional, top-down state-centric peacebuilding regimes requires alternative frameworks that can address the root causes of conflicts. This article analyses Hizmet transnational movem…

The failure of traditional, top-down state-centric peacebuilding regimes requires alternative frameworks that can address the root causes of conflicts. This article analyses Hizmet transnational movement's diagnostic model that envisages a symbiotic conflict triad of poverty, ignorance, and disunity. It contends that this tripartite model constitutes a critical analytical device which, in contrast to unidimensional thinking characteristic of theorization and practice in the field overarches these silos. By qualitatively analysing the movement's discourses of and praxis around gender with respect to educational institutions, dialogue work and humanitarian endeavour at a global level, this study shows how these three components mutually contribute to a vicious circle. The analysis shows that the strategic response of the movement is an integrated grassroots ecosystem aimed at creating a virtuous counter-cycle. Its education programmes not only address ethical ignorance but build intra and cross-group solidarity and economic access as the primary intervention. The paper argues that the Hizmet model offers a significant theoretical contribution in recharacterizing conflict as a complex multi-faceted systemic breakdown, and for peacebuilding strategies which specifically deconstruct its inter-related economic, epistemological, and socio-relational dynamics.

This record contains corrected file replacing the file associated with DOI 10.1016/j.aime.2024.100151 (https://zenodo.org/records/14039092). The previous version should no longer be used.

A Takayama–Judge spatial price-equilibrium model of the global crude oil market, solved daily over a 518-day horizon on a network of 2,662 oil fields, 537 refineries, 279 ports and 4,974 pipelin…

A Takayama–Judge spatial price-equilibrium model of the global crude oil market, solved daily over a 518-day horizon on a network of 2,662 oil fields, 537 refineries, 279 ports and 4,974 pipeline endpoints. Three nested scenarios: crude only (SC1), with a biomass-to-syncrude spot market (SC2), and with a forward-contract market for syncrude (SC3). Applied to a 2026 closure of the Strait of Hormuz and calibrated against 183 daily closes of Europe Brent Spot FOB from 2 January to 22 September 2026. The archive contains the model, the input data it needs, the reference results for every reported value, and the scripts that draw the paper's three figures, which are included. README.md gives the commands; RESULTS.md gives each reported value and the command that produces it; DATA_SOURCES.md gives the provenance of every input. Running a scenario needs Python 3.9 or later and the pinned packages in requirements.txt. A 518-day scenario takes 20 to 45 minutes.

Introduction: Older adults undergoing major emergency orthopedic surgery face high risks of postoperative delirium (POD), a debilitating complication. Inflammation and stress, reflected by the neutrop…

Introduction: Older adults undergoing major emergency orthopedic surgery face high risks of postoperative delirium (POD), a debilitating complication. Inflammation and stress, reflected by the neutrophil-to-lymphocyte ratio (NLR), may contribute to POD. This study evaluates NLR as a predictor of POD in this vulnerable cohort. Material and methods: This study employed a descriptive cross-sectional design at Shohada Hospital, Tabriz, Iran, with 88 elderly patients undergoing major emergency orthopedic surgery. Convenience sampling and Cochran’s formula determined sample size. Inclusion criteria focused on age (≥65), surgery type, preoperative CBC availability, and post-op assessment capability, while excluding pre-existing cognitive impairment or confounding conditions. Data on demographics, clinical factors, and NLR were collected, with postoperative delirium as the primary outcome. Results: A total of 88 patients were evaluated, with 26 (29.5%) developing postoperative delirium. Advanced age (P<0.001), higher ASA status (P=0.007), longer surgery (P=0.018) and delay times (P=0.005), lower albumin (P=0.001), and elevated NLR (P<0.001) were associated with delirium. Preoperative NLR (OR=1.89, 95% CI:1.29-2.77, P=0.001) independently predicted delirium. Conclusion: Preoperative NLR is a significant independent predictor of postoperative delirium in elderly patients undergoing major orthopedic surgery, suggesting its utility in risk stratification and early intervention strategies for this vulnerable population.

Актуальность. Острый пиелонефрит во время беременности представляет  Pсистемную инфекцию мочевых путей, способную сопровождаться тяжёлыми  материнскими осложнениями и ассоциироваться с небла…

Актуальность. Острый пиелонефрит во время беременности представляет  Pсистемную инфекцию мочевых путей, способную сопровождаться тяжёлыми  материнскими осложнениями и ассоциироваться с неблагоприятными  акушерскими исходами. Вместе с тем понятие «реабилитации после  пиелонефрита» остаётся недостаточно стандартизированным: доказательная  база значительно лучше разработана для лечения активной инфекции, скрининга бактериурии и наблюдения за рецидивами, чем для специальных  немедикаментозных или физиотерапевтических программ.