📢 #SpecialIssue AI-Based Combinatorial Optimization and Multi-Objective Optimization 📅31 December 2026 👨‍🔬Guest Editor: Dr. Miguel Angel Ortiz-Barrios from Universidad de la Costa, Colombia 🔗mdpi.com/journal/applsci/spe… #AI #combinatorialoptimization #multiobjectivedecisionmaking #metaheuristics #hybridintelligentsystems #machinelearning

Binary Latent Protein Fitness Landscapes for Quantum Annealing Optimization 1 This work introduces Q‑BioLat, a framework that projects protein sequences into a compact binary latent space and models fitness as a quadratic unconstrained binary optimization (QUBO) problem, enabling efficient combinatorial search and direct compatibility with quantum annealers. 2 Protein language models (ESM‑2) supply contextual embeddings; these are reduced in dimensionality via random projection or PCA, then binarized using median‑thresholding to generate balanced, interpretable latent codes that still capture biologically relevant variation. 3 A ridge‑regularized QUBO surrogate is trained on the binary codes, incorporating both unary and pairwise interactions. The resulting linear‑plus‑quadratic objective is immediately usable by simulated annealing, genetic algorithms, or hardware Ising solvers without further modification. 4 Experiments on the ProteinGym GFP benchmark show that simulated annealing and genetic algorithms consistently climb the surrogate landscape, and the nearest‑neighbor real sequences retrieved from optimized codes lie in the top ~90 % of the training fitness distribution, indicating successful navigation of high‑fitness regions. 5 Representation geometry proves crucial: PCA‑based latent codes outperform random projections in optimization performance, despite identical Spearman correlation, highlighting that the structure of the latent space, not just predictive accuracy, governs search efficacy. 6 Retrieval‑based decoding maps optimized binary codes back to actual protein sequences by Hamming‑distance search among training variants, providing a conservative, interpretable bridge between latent solutions and real proteins without training a generative decoder. 7 The framework scales with O(m²) surrogate parameters and search complexity, keeping the heavy protein‑language‑model inference fixed while allowing rapid, repeated optimization in the small binary latent space. 8 Future directions include learning binary representations jointly with the surrogate, expanding beyond pairwise terms, and deploying the QUBO model on quantum annealing hardware to accelerate protein engineering at scale. Code: github.com/HySonLab/Q-BIOLAT Paper: arxiv.org/abs/2603.17247 #ProteinEngineering #QuantumComputing #MachineLearning #Bioinformatics #ProteinDesign #QUBO #CombinatorialOptimization

A web-based Traveling Salesman Problem solver for experimenting with graphs up to 30 nodes earned a Proof of Usefulness score of 21 in HackerNoon’s contest. - hackernoon.com/building-a-we… #softwaredevelopment #combinatorialoptimization

Thrilled to announce ISCO 2026—the 9th International Symposium on Combinatorial Optimization—coming to Kuşadası, Türkiye, May 6–8, 2026. 🔗 isco2026.com.tr #CombinatorialOptimization #OperationsResearch #Algorithms #Optimization #GraphTheory #LNCS #Scopus #AcademicTwitter

Alongside our own presentations, we joined two inspiring trainings last week. It’s been both fun and productive to learn and exchange ideas with our consortium partners! 🌍🤝✨ #HorizonEurope #MSCA #UrbanProblems #CombinatorialOptimization

CSoNet 2013: 12 papers accepted. Authored by Weili Wu: 5 of them. 🤯Wow… what an extraordinary concentration of brilliance. Or is it something else entirely? @UTDCompSci @UTDJonsson @UT_Dallas @UMNews @IEEEorg @ComSoc @NSF @USEEOC #IEEEFellow #AcademicEthics #AcademicMisconduct #AcademicNepotism #ConferenceCodesofConduct #ConferenceEthics #TheoreticalComputerScience #COCOON #CombinatorialOptimization #AlgorithmDesign #FinancialFraud #GrantAbuse #MeTooSTEM #IEEE #ConflictOfInterest #FamilyCartels #TitleVII #NSFGrants #AcademicNepotism #SelfRewardSystem #ResearchIntegrity #WASA #CSoNet

1/11 Excited to present our latest work "Scalable Discrete Diffusion Samplers: Combinatorial Optimization and Statistical Physics" at #ICLR2025 on Fri 25 Apr at 10 am! #CombinatorialOptimization #StatisticalPhysics #DiffusionModels

🔥 Read our Review Paper 📚 A Systematic Review on Reinforcement Learning for Industrial Combinatorial Optimization Problems 🔗 mdpi.com/2076-3417/15/3/1211 👨‍🔬 by Mr. Miguel S. E. Martins et al. 🏫 @ULisboa_ #combinatorialoptimization #reinforcementlearning

ALERT: London Heathrow Airport closed bc fire. Below: incoming in-air #LHR (EGLL) flights. All diverted, e.g. to Amsterdam. We welcome them here in NL🇳🇱👍. Interesting planning ahead: crews, aircraft, passengers, luggage, hotels... #scheduling #combinatorialoptimization #math🤓

🚀 Highly viewed in MAKE: Bayesian Optimization Using Simulation-Based Multiple Information Sources over Combinatorial Structures 📄 Read here: mdpi.com/2504-4990/6/4/110 @ComSciMath_Mdpi #BayesianOptimization #CombinatorialOptimization

A recent research published @IEEEAccess from #TUS scientists enhanced the scalability of #IsingMachines, leading to greater applicability in real-world #CombinatorialOptimization problems.

🎊 We have cooperated with Dr. Jose Torres-Jimenez on the topic of "#CombinatorialOptimization for #Constructing Covering Arrays and #Sequence Covering Arrays" buff.ly/3ynWzgl Papers are welcomed before 31 August 2024! #MDPIOpenAccess #ComSciMath_Mdpi

UCD CS academics to co-organise prestigious @dagstuhl seminars - Professor @keanema on #ExplainableAI for #SequentialDecisionMaking and Assist. Prof. @Deepak_UCD on #MachineLearning Augmented Algorithms for #CombinatorialOptimization Problems. ucd.ie/cs/news/ucdcsacademic… #XAI

Dynamic and stochastic #Inventory #Routing problems (IRPs) still inspire you fear? Check how #CombinatorialOptimization problems with large state and action spaces can be handled with hybrid #MachineLearning and #Optimization pipelines: arxiv.org/abs/2402.04463 (1/5)

This is joint work with @sebaleh, @HochreiterSepp and @willberghammer . Join us at #NeurIPS2023 and also watch out for several other exciting works on #DeepLearning for #CombinatorialOptimization! 🌐 Link to Paper: arxiv.org/abs/2311.14156

It has been a pleausure to give the @MunichQuantum colloquium today on applications of #quantumcomputers in #machinelearning and #combinatorialoptimization. Thanks for the kind invitation.

#7AYW #Day3 #CombinatorialOptimization #MachineLearning Nuria Gómez-Vargas @justnuu_ shows a predict-and-optimize approach to guide the training of ML models with performances on the optimization problem and to enhance sparsity in the feature space for decisions explainability.

#7AYW #Day3 #CombinatorialOptimization #MachineLearning Léo Baty presents a policy for the dynamic VRPTW, which ranked first of the competition @EuroNeuripsVRP. It relies on a #DeepLearning pipeline with a prize collecting VRPTW combinatorial optimization layer.

#7AYW #Day1 #CombinatorialOptimization #MachineLearning Francesco Paolo Saccomanno proposes a reinforcement learning strategy to address #BinPacking, where the agent is trained to imitate a classic heuristic, the “best fit” strategy.

#7AYW #CombinatorialOptimization & #MachineLearning Antonio Consolo studies a variant of multivariate randomized regression trees and presents a decomposition training algorithm with a heuristic for the reassignment of the input vectors along the branching nodes of the tree.