Uncertainty Quantification of Quantum Chemical Methods - IPAM at UCLA

Explore uncertainty quantification in quantum chemical methods through this 42-minute conference talk presented by Markus Reiher from ETH Zurich at IPAM's Large-Scale Certified Numerical Methods in Quantum Mechanics Workshop. Delve into topics such as automation, kinetic modeling, transferability, discretization errors, benchmark results, and continuous benchmarking. Examine knowledge-based error estimation, Gaussian processes, reaction network exploration, and multiconfiguration DMRG. Discover entanglement measures, selection algorithms, and user interfaces for quantum chemical calculations. Learn about reference data, D3 correction, training data, and automated workflows in the context of quantum chemistry. Gain insights into the fundamental problems and challenges in quantifying uncertainties in quantum chemical methods, and understand the importance of error compensation and data-driven machine learning approaches in this field.

Introduction
Outline
Context
Automation
Camoton
Example
Network
Data
Kinetic modeling
Uncertainty quantification
Transferability
General Remarks
The Most Fundamental Problem
Discretization Errors
Individual Absolute Errors
Error Compensation
Benchmark Results
Continuous Benchmarking
Knowledge Based Error Estimation
Examples
Gaussian Processes
Reaction Network Exploration
Soap Kernel
Standard Database
Multiconfiguration
DMRG
Entanglement Measures
Selection Algorithm
User Interface
Reference Data
D3 Correction
Training Data
Technical Details
Data Machine Learning
Automated Workflow
Related Work
Conclusion
References