Motion and path planning are fundamental to intelligent robotic systems, enabling navigation. The objective is to generate collision-free trajectories in obstacle-rich configuration spaces (C-spaces) while meeting performance constraints. In environments with narrow passages planning becomes especially difficult, as feasible regions have low measure and are rarely reached by random sampling. Classical sampling-based planners are probabilistically complete but inefficient in such regions. Learning-based planners like MPNet offer fast inference but often produce infeasible paths in cluttered areas, requiring expensive postprocessing. To address this trade-off, we propose a hybrid framework that combines improved sampling, structural abstraction, and neural prediction. A modified bridge-test sampler applies directional perturbations and corridor checks to generate reliable narrow passage samples. These are clustered into a sparse set of representative bridge points, which serve as nodes in a global graph. At query time, a greedy heuristic search explores this graph, using a neural local segment generator to connect nodes. We validate the approach on 2D maze maps, 3D voxel environments, and a 12-DOF manipulator performing a plugging task inside a simulated nuclear steam generator. Across all tasks, our method significantly outperforms classical and learning-based baselines in terms of success rate and planning time in narrow-passage-dominated scenarios. The inclusion of the repair module, under relaxed assumptions, also allows the framework to retain a generalized form of probabilistic completeness.

Efficient greedy optimization method for k-means

Dynamic distributed production scheduling problem with simultaneous job arrivals and machine breakdowns: Learning to learn iterated greedy algorithm

A two-stage iterated greedy algorithm for distributed blocking flowshop scheduling problem

With the rapid advancement of communication technology, many time-sensitive applications have emerged, necessitating real-time networks for efficient data transmission. Time-Sensitive Network (TSN) is pivotal in supporting time-critical operations and is widely acknowledged as the optimal solution for meeting real-time network requirements. However, TSN requires effective routing and scheduling strategies to achieve robust real-time performance and bounded low latency. Therefore, this paper proposes a joint routing and scheduling (JRS) method based on a greedy algorithm to address the scheduling challenges associated with time-sensitive traffic in TSN. Specifically, we introduce a bandwidth-balanced routing algorithm to enhance problem-solving success rates while simultaneously reducing node load within the network. Additionally, we present a greedy-based scheduling algorithm that rapidly generates feasible solutions for network scheduling within short time frames. Integrating these routing and scheduling methods expands the solution space for this problem domain. Finally, experimental results demonstrate that our proposed method effectively caters to simple and complex environments encountered in practical use cases with broad applicability. This approach exhibits minimal execution time while maintaining high feasibility in typical topologies; furthermore, it holds the potential to adapt to multiple optimization algorithms.

In recent years, people have increasingly sought to generate exercise trajectories that embody specific semantic shapes in order to create GPS art and share it on social platforms. This trend has created an urgent demand for navigation paths with specific semantic meanings on smartwatches and smartphones. Current methods mainly rely on manual design and lack efficient automation. Therefore, this study proposes a novel method for automatically obtaining navigation paths with specified shapes by retrieving graphics similar to the input graphic shape from the road network. This method uses invariant spatial relationships, such as turning angles and length ratios, along with graph matching techniques to establish one-to-one or one-to-many correspondences between line segments in the input individual graphics and those in the road network. This enables the retrieval of individual graphics within the road network. Based on this, a greedy strategy-based algorithm is proposed to solve the combined graphics retrieval problem. The results are evaluated to ensure high quality. The accuracy and effectiveness of our method are validated through experimental results using simulated and real road network data from five different regions. Furthermore, shape-constrained graphics retrieval expands the application domain of spatial scene matching.

Enhancing recommender systems: the improved serendipity-oriented greedy (ISOG) algorithm for balanced accuracy, serendipity, and efficiency

Abstract PI3K and AKT pathway inhibitors Alpelisib, Inavolisib, and Capivasertib (PI3K/AKTi) are FDA approved treatments with demonstrated significant clinical benefit, improving progression free survival for Hormone receptor positive (HR+) mBC. Hyperglycemia can commonly develop from these agents. While knowledge about disparities in outcomes based on race and comorbidities in certain mBC subtypes is emerging, real world data about PI3K/AKTi is limited. In this propensity score-matched cohort study we used TriNetX de-identified EHR data from multiple health systems to compare mortality of patients with breast cancer diagnoses treated with PI3K/AKTi. We identified 2372 patients, female (F) and male (M), stratified in African American (AA; n= 262F, 8M), White (W; n=1933 F, 25M) and Asian (A, n=144F) cohorts. Qualification into the three race-based cohorts required presence of a C50 ICD-10-CM diagnosis code and treatment with PI3K/AKTi. Mortality outcomes were also compared in separate cohorts examining hazard ratios for conditions developed before or after treatment: diabetes (E08-E13), hypertension (I10-I15), obesity (E66), and hyperlipidemia (E78.4-5). Cohorts were matched for age, comorbidities, and treatment lines using 1:1 matching with a greedy nearest neighbor search. All-time Odds ratios (OR) and/or Hazard ratios (HR) with 95% confidence intervals are reported as an effect size and significance estimation. Compared to non-diabetic patients, those diagnosed with diabetes after treatment had higher odds of mortality post-matching (OR = 1.53, 95% CI 1.19-1.97), but those with pre-existing diabetes showed no significant difference in mortality (matched OR = 1.19, 95% CI 0.97-1.45). Both pre-existing hypertension and hypertension diagnosed post-treatment were associated with higher mortality compared to non-hypertensive patients: matched ORs = 1.23 (95% CI 1.02-1.47) and 2.02 (95% CI 1.36-3.00), matched HRs = 1.204 (95% CI 1.05-1.38) and 1.911 (95% CI 1.42-2.57), respectively. No statistically significant difference in mortality odds was seen comparing the AA cohort to W cohort (matched OR = 0.99, 95% CI 0.70-1.39) or A cohort to W cohort post matching (OR = 0.72, 95% CI 0.45-1.16). New-onset hyperlipidemia was not statistically significant after matching (OR = 1.15, 95% CI 0.91-1.45; HR = 1.15, 95% CI 0.97-1.35). No significant association was seen with obesity before or after treatment (all CI including 1). This study demonstrates that hypertension diagnosed before and after treatment with PI3K/AKTi is associated with higher mortality, as is diabetes diagnosis post-treatment. We did not identify disparities in mortality between our 3 racial cohorts. Future studies are needed to understand interactions between individual PI3K/AKTi agents and comorbidities to inform appropriate management strategies and ultimately improve patient outcomes. Citation Format: T. S. Morales, A. Strong, J. Petucci, M. K. Vasekar. Investigating association of comorbidities and race with all-cause mortality outcomes of PI3K inhibitor use in metastatic breast cancer (mBC) [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS5-06-25.

The integration of Unmanned Aerial Vehicles (UAVs) with the Internet of Things (IoT) is revolutionizing a wide range of applications. However, collecting massive sensing data from large-scale fields efficiently remains challenging, constrained by the limited energy of UAVs and sensing nodes. Existing schemes lack the computational intelligence of an Edge Server (ES) for deep coordination. To address this, we propose DC-CSAP, a novel “Edge-UAV-End” collaborative data collection framework. DC-CSAP introduces a systematic workflow orchestrated by the ES, which is operationalized through four dedicated collaboration mechanisms: (1) In our ES–UAV collaboration, we devise a two-phase path optimization algorithm that hybridizes Simulated Annealing (SA) with a convex-hull-inspired greedy method. (2) The ES–ISN collaboration features a prediction-based binary vector mechanism, transmitting only inaccurate data to slash communication overheads. (3) The UAV–ISN and (4) Inter-ISN protocols ensure efficient data exchange and aggregation. Extensive simulations validate that DC-CSAP outperforms benchmarks in terms of Correct Prediction Rate (CPR), energy efficiency, and UAV path length.

A Greedy Genetic Hybrid Heuristic for Solving the Open-pit Production Scheduling Problem

This paper introduces LeapNP (Learning and Planning Framework for Numeric Problems), a lightweight, Python-native framework engineered to support both classical and numeric planning tasks. Designed with a fully modular interface, it specifically aims to facilitate the seamless integration of deep learning methodologies. The design philosophy of LeapNP stems from the observation that traditional planners, while highly efficient, lack the necessary flexibility for experimental research, particularly at the intersection of learning and planning. Most state-of-the-art engines are built as highly optimized, rigid executables that are resistant to internal modification. LeapNP disrupts this paradigm by offering a framework where the entire planning stack is accessible and mutable. Users can seamlessly plug in custom implementations for grounding, define novel state representations, or design bespoke search strategies, thereby enabling a level of integration with learning models that is currently impractical with standard tools. By significantly lowering the engineering barrier, our planner fosters rapid experimentation and accelerates research in neuro-symbolic planning. We also present a comprehensive suite of search algorithms, designed to evaluate different properties of learned heuristics. These include two algorithms designed to exploit batching to maximize inference throughput, and a greedy algorithm meant to test the intrinsic robustness of the learned models, running them as general policies.

Laser interstitial thermal therapy (LITT) has emerged as a minimally invasive alternative to open craniotomy for patients deemed unsuitable for surgery due to deep-seated or eloquent lesion location, age, frailty, or comorbidities. However, its use in newly diagnosed deep-seated low-grade glioma (nLGG) has not been elucidated. We aimed to evaluate the safety and efficacy of LITT for deep-seated nLGG compared with a similar surgical cohort. We retrospectively reviewed patients with unifocal, deep-seated nLGG treated with either LITT or surgical resection between 2013 and 2024. Demographic, perioperative, and follow-up data were compared between groups. Kaplan-Meier assessed progression-free and overall survival outcomes. To address baseline tumor volume differences, a subset analysis was performed using a greedy nearest-neighbor algorithm to generate a 1:1 matched cohort based on tumor volume. A total of 15 patients in the study group (median age 46 [IQR: 34-53] years, 40.0% men) were compared with 51 patients (median age 38 [IQR: 29-54] years, 43.1% men) in the control group. There were no significant differences in in-hospital complications (P = .999), 30-day complications (P = .999), or complications between 30 days and 3 months (P = .713), new postoperative motor or speech deficits (0.999) between groups. Postoperative adjuvant chemotherapy (23.1% vs 46.9%, P = .217) and radiation (23.1% vs 44.7%, P = .210) rates did not differ significantly. Among high-risk patients, time to adjuvant chemotherapy (64.7 vs 77.7 days) and radiation (36.0 vs 53.6 days) was earlier in the LITT group, although not statistically significant. Kaplan-Meier analysis showed no statistically significant differences in progression-free survival or overall survival between groups. On matched pair analysis, there remained to be no statistically significant differences in outcomes observed between LITT and craniotomy groups. This pilot study is the first to suggest that LITT is a safe treatment option for patients with deep-seated nLGG, offering comparable outcomes with surgical resection.

The combinatorial explosion inherent to multi-component systems limits their experimental exploration and ultimately chemical discovery. Here, we introduce a statistics-based group-testing strategy, which we couple with luminescence quenching assays to efficiently identify cooperative molecular interactions. Utilizing the quenching of a photosensitizer as a quick readout for chemical activity, 4,950 substrate pairs were screened in only 504 experiments, enabled through a combinatorial design theory-based pooling approach and iterative deconvolution. Therefore, two algorithms-a greedy algorithm for group design and an iterative sectioning deconvolution method to resolve active pairs-were implemented. Fifteen cooperative pairs were identified, and the nature of their interactions and the resulting electronic perturbations were investigated. In a systematic follow-up screen, it was found that the identified active pairs exhibit high reactivity towards a broad group of reaction partners. One such pair led to the discovery of a bench-stable reagent, enabling efficient and regioselective trifluoromethylthiolation reactions. This work establishes a broadly applicable framework for accelerating the discovery of cooperative reactivity through optimized experimental designs.

Purpose This study aims to develop a novel hybrid Genetic Algorithm (GA) for Flexible Job-shop Scheduling Problems (FJSP) to improve makespan. Design/methodology/approach DIGA, a Reinforcement Learning (RL)-driven algorithmic framework is proposed. This algorithmic framework consists of GA, Iterated Greedy Algorithm (IGA) and Double Q-Learning Algorithm (DQLA), which is an RL-based technique. IGA was used as the local search technique to improve the performance of GA. The DQLA tuned GA parameters – selection strategy, crossover rate and mutation rate – to their proper values so that it resulted in better makespan. Brandimarte benchmark instances were solved using this algorithm and statistical tests were carried out to assess its performance. Comparison with results from recent literature was also made. Findings Results show that DIGA is a statistically significant, stable, consistent and robust algorithm that can be used to solve FJSP. The tournament selection strategy was chosen by DQLA on vast majority of occasions, followed by Roulette wheel and Ranking strategies. The presence of IGA as local search technique has considerably reduced the number of crossover and mutation operations needed, helping faster convergence to the best makespan. Originality/value Tuning of all three GA parameters for FJSP was never reported in literature. Proposed DIGA is a novel algorithmic framework.

Outcomes associated with distal femur fractures treated with distal femur replacement compared to open reduction internal fixation in elderly patients.

A novel elite-preserving iterated greedy algorithm with Q-learning for cascaded flowshop joint scheduling problem

Server deployment is a fundamental task in mobile edge computing: where to place the edge servers and what user cells to assign to them. To make this decision is context-specific, but common goals are (1) computing efficiency: maximize the amount of workload processed by the edge, and (2) communication efficiency: minimize the communication cost between the cells and their assigned servers. We focus on practical scenarios where the user workload in each cell is unknown and time-varying, and so are the effective capacities of the servers. Our research problem is to choose a subset of candidate servers and assign them to the user cells such that the above goals are sustainably achieved under the above uncertainties. We formulate this problem as a stochastic bilevel optimization, which is strongly NP-hard and unseen in the literature. By approximating the objective function with submodular functions, we can utilize state-of-the-art greedy algorithms for submodular maximization to effectively solve our problem. We evaluate the proposed algorithm using real-world data, showing its superiority to alternative methods; the improvement can be as high as 55%.

Achieving environmental sustainability in AI systems requires algorithms that balance computational efficiency with energy conservation. Best-First Search (BFS), a foundational heuristic-based graph traversal technique, plays a pivotal role in areas such as robotics, automated planning, and symbolic reasoning. This work presents a comparative evaluation of five recent advances in BFS, highlighting innovations in parallelism, heuristic learning, and integration with large language models (LLM). Building on these insights, we propose ECO-GH (Energy Constrained Opportunistic Greedy Heuristic Search), a novel variant of BFS tailored for climate-aware computation. ECO-GH incorporates energy-sensitive tie-breaking, opportunistic node expansion, adaptive parallelism, and dynamic heuristic switching. The experimental results on standard planning benchmarks reveal that ECO-GH achieves up to 34% lower energy consumption and a greater 25% reduction in runtime compared to state-of-the-art approaches, without compromising the quality of the solution. These findings position ECO-GH as a scalable and sustainable search framework for environmentally responsible applications, including renewable energy planning, autonomous electric mobility, and intelligent infrastructure systems.

Abstract In the realm of robust optimization the k -adaptability approach is one promising method to derive approximate solutions for two-stage robust optimization problems. Instead of allowing all possible second-stage decisions, the k -adaptability approach aims at calculating a limited set of k such decisions already in the first-stage before the uncertainty is revealed. The parameter k can be adjusted to control the quality of the approximation. However, not much is known on how many solutions k are needed to achieve an optimal solution for the two-stage robust problem. In this work we derive bounds on k which guarantee optimality for general non-linear problems with integer decisions where the uncertainty appears in the objective function or in the constraints. For convex uncertainty sets we show that for objective uncertainty the bound depends linearly on the dimension of the uncertainty, while for constraint uncertainty the dependence can be exponential, still providing the first generic bound for a wide class of problems. Additionally, we provide approximation guarantees if k is smaller than the derived bounds. The results give new insights on how many solutions are needed for problems as the decision dependent information discovery problem or the capital budgeting problem with constraint uncertainty. Finally, for finite uncertainty sets we show that calculating the minimal k for which k -adaptable and two-stage problems are equivalent is NP-hard and derive a greedy method which approximates this k for the case where no first-stage decisions exist.

This paper focuses on the target Q-coverage and sensor deployment problem, where the coverage requirement of each target is different. Previous studies in the literature solved the Q-coverage optimization as a single-objective problem considering random sensor deployment. In this work, Q-coverage problem in directional sensor network is solved as a multi-objective optimization problem, where objectives are the maximization of the overall target coverage and the minimization of the number of active sensors. Maintaining their generic structures the same, three existing and well-known multi-objective genetic algorithms (MOGAs): strength Pareto evolutionary algorithm 2 (SPEA2), nondominated sorting genetic algorithm II (NSGA-II), and multi-objective evolutionary algorithm based on decomposition (MOEA/D) are modified to solve the proposed multi-objective Q-coverage problem. Since the sensor positions can significantly improve the overall coverage, two different sensor deployment algorithms are proposed to find the suitable positions of sensors, one is heuristic, and the other is based on particle swarm optimization. The MOGAs determine the optimal orientations of the sensors. For the implementation of the modified MOGAs, a new mutation operator, compatible for implementing the problem, has been designed. The impact of five different network parameters: the number of targets, the number of sensors, the number of orientations, the sensing radius values, and the coverage requirement, on the two objectives are analyzed. The performances of the three modified MOGAs are compared based on three performance metrics, hypervolume (HV), inverted generational distance (IGD), and spread. To analyze the robustness of the modified MOGAs, the sensitivity analysis is performed. In addition, the performances of the three modified MOGAs are compared with a genetic algorithm and a greedy algorithm, existing in the literature. Experimental results show that the modified MOGAs need, on an average, 16.28% fewer sensors than the existing genetic algorithm and 2.9% fewer sensors than the existing greedy algorithm, while achieving 7.8% and 1.18% higher target coverage, respectively. To show the scalability and effectiveness of the proposed MOGAs, they are executed on both a large scale network and a real network. Finally, the results are validated using the Wilcoxon signed rank test based on the performance metrics.