←Back to feed
🧠 AI⚪ Neutral
Adaptive Uncertainty-Guided Surrogates for Efficient phase field Modeling of Dendritic Solidification
arXiv – CS AI|Eider Garate-Perez, Kerman L\'opez de Calle-Etxabe, Oihana Garcia, Borja Calvo, Meritxell G\'omez-Omella, Jon Lambarri||1 views
🤖AI Summary
Researchers developed a new AI-powered surrogate model using XGBoost and CNNs to significantly reduce computational costs in phase field simulations for metal solidification processes. The adaptive uncertainty-guided approach achieves accurate predictions while requiring fewer expensive simulations and reducing CO2 emissions in additive manufacturing applications.
Key Takeaways
- →New surrogate model combines XGBoost and CNNs with uncertainty-driven adaptive sampling to reduce computational costs in phase field simulations.
- →The approach uses Monte Carlo dropout for CNNs and bagging for XGBoost to identify high-uncertainty regions requiring additional samples.
- →Framework outperforms traditional Optimal Latin Hypercube Sampling optimized via Particle Swarm Optimization in efficiency.
- →Research addresses critical microstructural control challenges in additive manufacturing of metals.
- →Study includes environmental impact assessment by measuring associated CO2 emissions alongside computational performance.
#machine-learning#xgboost#cnn#computational-modeling#additive-manufacturing#uncertainty-quantification#surrogate-models#phase-field#materials-science
Read Original →via arXiv – CS AI
Act on this with AI
Stay ahead of the market.
Connect your wallet to an AI agent. It reads balances, proposes swaps and bridges across 15 chains — you keep full control of your keys.
Related Articles