Academic profile · Research & publications

Jose L.
Salmeron

Full Professor of Artificial Intelligence (Catedrático)

Three decades of research at the frontier of artificial intelligence — 150+ ISI-indexed publications, 7,000+ citations, and a top-2% global ranking. This site collects my academic work: research lines, funded projects and papers.

Looking for my industry & executive profile? joselsalmeron.github.io →

Explore research
Portrait of Prof. Jose Luis Salmeron
Stanford Top 2% Scientist
0 h-index
0 Citations
0 ISI Publications
0 Funding as PI
Top 2% Global Scientist
01 — About

Research at the frontier of AI

I am Jose L. Salmeron, a researcher who has spent three decades advancing artificial intelligence — from foundational theory to methods that hold up against the messiness of real-world data. I hold the rank of Catedrático, Spain's highest professorial rank, currently at CUNEF Universidad (Madrid) after a decade as full professor at Universidad Pablo de Olavide, with ongoing international appointments across Europe and Latin America.

My research lives at the intersection of explainable AI, federated & privacy-preserving learning, quantum computing, causal machine learning and neuro-symbolic AI — much of it grounded in fuzzy cognitive maps and applied to healthcare, finance and critical infrastructure. Over 150 papers indexed in ISI Web of Science, 7,000+ citations and an h-index of 39 place me among the top 2% of scientists worldwide in the Stanford/ISI ranking.

Beyond publishing, I have led research programmes as Principal Investigator with €20M+ in competitive funding from EU, national (CDTI) and international funders, and collaborate with groups such as McGill University and the Canadian CIHR. I supervise doctoral researchers, serve on editorial boards and scientific committees, review for leading journals, and regularly deliver keynotes — working to keep frontier AI both rigorous and genuinely useful.

02 — Research

Interests & expertise

Frontier methods, deployed responsibly.

🔍

Explainable AI (XAI)

Post-hoc and intrinsic interpretability for audit-ready decisions in regulated, high-stakes domains.

🌐

Federated Learning

Privacy-preserving distributed learning across institutions without centralising sensitive data.

⚛️

Quantum Computing

Variational quantum circuits and quantum-inspired AI for scarce-data clinical decision support.

🧬

Causal Machine Learning

Non-monotonic causal discovery and reasoning beyond correlation, with Kolmogorov–Arnold methods.

🪞

Digital Twins

Virtual replicas for cardiovascular risk and process modelling from limited datasets.

🧩

NeuroSymbolic AI

Bridging neural learning and symbolic reasoning for interpretable, knowledge-grounded models.

🎭

Generative Adversarial Networks

Divergence-driven GANs for semi-supervised anomaly detection and synthetic data generation.

🎯

Recommender Systems

Scalable, model-agnostic explainability for recommendation at production scale.

03 — Career

Academic appointments

This is my academic track only. For industry & executive roles, see joselsalmeron.github.io.

  1. 2022 — present

    Professor (Catedrático) of Computer Science & AI

    CUNEF Universidad · Madrid, Spain

  2. 2012 — 2022

    Professor (Catedrático) of Artificial Intelligence

    Universidad Pablo de Olavide · Seville, Spain

  3. 2003 — 2011

    Associate Professor of Information Systems & AI

    Universidad Pablo de Olavide · Seville, Spain

  4. 2014 — present

    Senior Associate Researcher of AI

    Universidad Autónoma de Chile · Chile

  5. 2015 — 2019

    Research Professor of Artificial Intelligence

    Univerzita Hradec Králové · Czech Republic

  6. 2015 — 2017

    Principal Investigator — Int. Network in Grey Systems

    The Leverhulme Trust · coord. De Montfort University

  7. 2013

    Visiting Scientist — Cybernetics & AI

    Technical University of Košice · Slovak Republic

  8. 2012

    Visiting Scientist — Systems & Industrial Engineering

    The University of Arizona · USA

  9. 2001

    Visiting Scholar — Information Systems & Quantitative Sciences

    Texas Tech University · USA

  10. 1997 — 2000

    Adjunct Professor (Profesor Asociado)

    Universidad de Huelva · Spain

  11. 1996 — 1997

    Teaching Assistant of Information Systems

    Universidad de Sevilla · Spain

04 — Projects

Research projects & programmes

The full record of applied and funded R&D I have led — €20M+ secured as Principal Investigator across EU, national and international funders.

05 — Open Source

Open-source software & models

Research-grade tools and models I have authored and released to the community.

TurboSwarm Rust · Python

Particle swarm optimization library

A full-featured particle swarm optimization library with a high-performance Rust core and a Python API. It supports multiple velocity rules (inertia, constriction, FIPS), swarm topologies, continuous/integer/binary/mixed search spaces, multi-objective optimization (MOPSO), constraint handling and parallel evaluation. It also extends PSO to optimise directly over grey (interval) numbers via a center–spread encoding, with convergence and swarm visualization throughout.

OptimizationPSOMulti-objectiveRust
Documentation →
swarmstate Rust · Python

State & checkpointing backend for agent orchestration

A framework-agnostic state management and checkpointing backend for agent orchestration frameworks such as LangGraph and CrewAI. A high-performance Rust core delivers fast serialization, O(1) immutable snapshots and incremental diffs — releasing Python's GIL on critical paths — while a stable, language-neutral MessagePack format lets you switch frameworks without losing state. It also resolves deterministic, rule-based routing natively in Rust to cut unnecessary LLM token use, with memory, Redis and disk backends.

Agent OrchestrationState ManagementCheckpointingRust
Documentation →
TinyLlama · Mac GGUF · ~700 MB

TinyLlama 1.1B Chat — Mac-quantized

A quantized release of TinyLlama-1.1B-Chat-v1.0 in GGUF format, optimised for Apple Silicon and consumer hardware. Uses Q4_K_M 4-bit importance-matrix (imatrix) quantization to preserve quality while shrinking the model for fully local inference via llama.cpp.

LLMQuantizationApple SiliconOn-device
View on Hugging Face →
fuzzytool Python · NumPy

Fuzzy inference systems toolkit

A fuzzy-logic toolkit for building fuzzy inference systems with a clean, composable API. Rules use natural logical operators (& | ~), membership functions, connectives and defuzzifiers are pluggable through small Protocol interfaces, and both Mamdani and Takagi–Sugeno (TSK) inference are supported. Pure Python, with built-in visualization — a modern alternative to scikit-fuzzy's control API.

Fuzzy LogicInference SystemsMamdaniTSK
Documentation →
GraphFinder Rust · Python

Graph traversal & pathfinding toolkit

An educational library for graph search and pathfinding, covering both uninformed (BFS, DFS, Dijkstra) and informed (Greedy, A*, Weighted A*) strategies under a single unified loop that differs only in frontier management and priority. It supports grid/maze worlds, weighted and implicit graphs, with search animations, frontier analysis and reproducible, seeded runs. A high-performance Rust core with a clean Python API.

PathfindingA* SearchAlgorithmsRust
Documentation →
06 — Publications

Publications

The full record from my academic CV — filter by type or search by title, venue or co-author. Live metrics on Google Scholar →

07 — Collaboration

Research collaborations

European UnionGovernment of CanadaCDTI · Spanish Ministry of ScienceThe Leverhulme TrustMcGill UniversityUniversidad Autónoma de ChileIntelAirbus
BBVAAccentureCapgeminiAltranVodafoneCiscoMicrosoftGilead
08 — Teaching

Teaching resources

Course platforms and materials for my students. Each course is protected by its own password — enter the one I share with your class.

09 — Contact

Let's collaborate

Open to research collaboration and co-authorship, consulting, EU & national R&D projects, scientific committees and academic keynotes.

salmeron@acm.org LinkedIn Google Scholar