Antonio del Rio Chanona

📢 𝗡𝗲𝘄 𝗽𝗮𝗽𝗲𝗿 alert! We document how 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜 𝗶𝘀 𝗿𝗲𝘀𝗵𝗮𝗽𝗶𝗻𝗴 𝗱𝗲𝗺𝗮𝗻𝗱 𝗳𝗼𝗿 𝗳𝗿𝗲𝗲𝗹𝗮𝗻𝗰𝗲𝗿 𝘀𝗸𝗶𝗹𝗹𝘀. 💼 💻

📊 We analysed 𝟯𝗠+ 𝗷𝗼𝗯 𝗽𝗼𝘀𝘁𝘀, used embeddings to cluster them into 𝟭𝟬𝟬+ 𝗳𝗶𝗻𝗲-𝗴𝗿𝗮𝗶𝗻𝗲𝗱 𝘀𝗸𝗶𝗹𝗹𝘀, and LLMs to classify them as s𝘂𝗯𝘀𝘁𝗶𝘁𝘂𝘁𝗮𝗯𝗹𝗲, 𝗰𝗼𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗮𝗿𝘆, or 𝘂𝗻𝗮𝗳𝗳𝗲𝗰𝘁𝗲𝗱.

🚨 New paper alert! Economic Agent Based Models have become increasingly data-driven. What does that mean and what impact can this have? arxiv.org/abs/2412.16591 📖✨

Chapter with @marcopangallo.bsky.social forthcoming in @sfiscience.bsky.social volume #Economics #ComplexSystems Part IV

👇🧵

Multi-Agent Data-driven Economic Conference in Venice

www.unive.it/web/en/8683/...

#Economics #ComplexSystems

Investigating the reliability and interpretability of machine learning frameworks for chemical retrosynthesis Machine learning models for chemical retrosynthesis have attracted substantial interest in recent years. Unaddressed challenges, particularly the absence of robust evaluation metrics for performance c...

𝗥𝗲𝗮𝗰𝘁𝗶𝗼𝗻 𝗣𝗹𝗮𝗻𝗻𝗶𝗻𝗴 𝘄𝗶𝘁𝗵 𝗗𝗲𝗲𝗽 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 ⚛️:

➡️ Machine learning frameworks for chemical retrosynthesis: pubs.rsc.org/en/content/a...

Foundation Models at Work: Fine-Tuning for Fairness in Algorithmic Hiring Foundation models require fine-tuning to ensure their generative outputs align with intended results for specific tasks. Automating this fine-tuning process is challenging, as it typically needs human...

𝗦𝗰𝗶𝗲𝗻𝘁𝗶𝗳𝗶𝗰 𝗟𝗟𝗠𝘀 🧠:

➡️ Foundation Models at Work: Fine-Tuning for Fairness in Algorithmic Hiring: arxiv.org/abs/2501.07324

➡️ Improved Scholarly Document Comprehension for Large Language Models: aclanthology.org/2024.sdp-1.28/

𝗗𝗮𝘁𝗮-𝗗𝗿𝗶𝘃𝗲𝗻 𝗢𝗽𝘁𝗶𝗺𝗶𝘇𝗮𝘁𝗶𝗼𝗻 ⛰️:

➡️ Surrogate-Based Optimization: arxiv.org/abs/2412.13948

➡️ Hierarchical planning-scheduling-control: arxiv.org/abs/2310.07870

➡️ Discrete and mixed-variable experimental design with surrogate-based approach: pubs.rsc.org/en/Content/A...

The automated discovery of kinetic rate models – methodological frameworks The industrialization of catalytic processes requires reliable kinetic models for their design, optimization and control. Mechanistic models require significant domain knowledge, while data-driven and...

𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗲𝗱 𝗠𝗼𝗱𝗲𝗹 𝗖𝗼𝗻𝘀𝘁𝗿𝘂𝗰𝘁𝗶𝗼𝗻 𝗮𝗻𝗱 𝗗𝗶𝘀𝗰𝗼𝘃𝗲𝗿𝘆 🔬:

➡️ The automated discovery of kinetic rate models: pubs.rsc.org/en/content/a...

➡️ Simplest Mechanism Builder Algorithm (SiMBA): arxiv.org/abs/2410.21205

Machine learning-assisted discovery of flow reactor designs - Nature Chemical Engineering Identifying the optimal geometry of continuous flow reactors is a major challenge due to the large available parameter design space. Here the authors combine a machine learning-assisted methodology wi...

#𝗕𝗮𝘆𝗲𝘀𝗶𝗮𝗻𝗢𝗽𝘁𝗶𝗺𝗶𝘇𝗮𝘁𝗶𝗼𝗻 & 𝗛𝘂𝗺𝗮𝗻-𝗶𝗻-𝘁𝗵𝗲-𝗹𝗼𝗼𝗽 📈

➡️ Bayesian optimization for discovery of reactor designs: www.nature.com/articles/s44...

➡️ Human-algorithm collaborative Bayesian optimization: arxiv.org/abs/2404.10949

Control-Informed Reinforcement Learning for Chemical Processes This work proposes a control-informed reinforcement learning (CIRL) framework that integrates proportional-integral-derivative (PID) control components into the architecture of deep reinforcement lear...

#𝗥𝗲𝗶𝗻𝗳𝗼𝗿𝗰𝗲𝗺𝗲𝗻𝘁𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 🤖:

➡️ Control-Informed Reinforcement Learning: arxiv.org/abs/2408.13566

➡️ Graph Neural Networks and Multi-Agent Reinforcement Learning: arxiv.org/abs/2410.18631

➡️ PC-Gym: Benchmark Environments for Process Control Problems: arxiv.org/html/2410.22...

Hello Bluesky! Our group @imperialchemeng.bsky.social researches #AI & #ChemicalEngineering. if you are interested, some of our 2024 research is below on:

#ReinforcementLearning 🤖
#BayesianOptimization 📈
#AutomatedModelDiscovery 🔬
#DataDrivenOptimization ⛰️
#ScientificLLMs 🧠
#ReactionPlanning ⚛️

Hello, Bluesky community! ️‍🔥

I’m a researcher in optimization and machine learning applied to chemical engineering.

Excited to connect, share ideas and science in this fresh space! 🚀