Publications related to Path Integral Methods in Atomistic Modelling

Genetic Optimization of Training Sets for Improved Machine Learning Models of Molecular Properties

Anatole von Lilienfeld, Nicholas John Browning

The training of molecular models of quantum mechanical properties based on statistical machine learning requires large data sets which exemplify the map from chemical structure to molecular property. Intelligent a priori selection of training examples is o ...

2017

Geometric Integrators for Nonadiabatic Molecular Quantum Dynamics Induced by the Interaction with the Electromagnetic Field

Aurélien Kevin Fabrice Patoz

Understanding the elementary steps involved in a chemical reaction forms the cornerstone of physical chemistry research. One way to deepen this understanding is by studying chemical and physical processes using linear and nonlinear spectroscopic techniques ...

EPFL2017

Safe Adaptive Importance Sampling

Martin Jaggi, Sebastian Urban Stich, Anant Raj

Importance sampling has become an indispensable strategy to speed up optimization algorithms for large-scale applications. Improved adaptive variants - using importance values defined by the complete gradient information which changes during optimization - ...

2017

Sampling at unknown locations, with an application in surface retrieval

Martin Vetterli, Benjamin Bejar Haro, Adam James Scholefield, Michalina Wanda Pacholska

We consider the problem of sampling at unknown locations. We prove that, in this setting, if we take arbitrarily many samples of a polynomial or real bandlimited signal, it is possible to find another function in the same class, arbitrarily far away from t ...

2017

On the performance of random reshuffling in stochastic learning

Ali H. Sayed, Stefan Vlaski, Bicheng Ying, Kun Yuan

In empirical risk optimization, it has been observed that gradient descent implementations that rely on random reshuffling of the data achieve better performance than implementations that rely on sampling the data randomly and independently of each other. ...

2017

Discrete least-squares approximations over optimized downward closed polynomial spaces in arbitrary dimension

Fabio Nobile, Giovanni Migliorati

We analyze the accuracy of the discrete least-squares approximation of a function

u

in multivariate polynomial spaces

P_\Lambda:=span\{y\mapsto y^\nu \,: \, \nu\in \Lambda\}

with

\Lambda\subset N_0^d

over the domain

\Gamma:=[-1,1]^d

, based on the s ...

2017

Classical Force Fields Tailored for QM Applications: Is It Really a Feasible Strategy?

Oliviero Andreussi

Classical molecular dynamics is more and more often coupled to quantum mechanical based techniques as a statistical tool to sample configurations of molecular systems embedded in complex environments. Nonetheless, the classical potentials describing the mo ...

Amer Chemical Soc2017

Learning-based subsampling

Volkan Cevher, Jonathan Mark Scarlett, Ilija Bogunovic, Yen-Huan Li, Luca Baldassarre, Baran Gözcü

The present invention concerns a method of sampling a test signal. The method comprises: acquiring (21) training signals sampled at a plurality of sampling locations; running (23) an optimization procedure for determining an index set of n indices, represe ...

2017

Sampling and Exact Reconstruction of Pulses with Variable Width

Martin Vetterli, Adam James Scholefield, Loïc Arnaud Baboulaz, Gilles Baechler

Recent sampling results enable the reconstruction of signals composed of streams of fixed-shaped pulses. These results have found applications in topics as varied as channel estimation, biomedical imaging and radio astronomy. However, in many real signals, ...

Institute of Electrical and Electronics Engineers2017

Just Sort It! A Simple and Effective Approach to Active Preference Learning

Matthias Grossglauser, Lucas Maystre

We address the problem of learning a ranking by using adaptively chosen pairwise comparisons. Our goal is to recover the ranking accurately but to sample the comparisons sparingly. If all comparison outcomes are consistent with the ranking, the optimal sol ...

2017

Path Integral Methods in Atomistic Modelling

Graph Chatbot

Chat with Graph Search

Genetic Optimization of Training Sets for Improved Machine Learning Models of Molecular Properties

Geometric Integrators for Nonadiabatic Molecular Quantum Dynamics Induced by the Interaction with the Electromagnetic Field

Safe Adaptive Importance Sampling

Sampling at unknown locations, with an application in surface retrieval

On the performance of random reshuffling in stochastic learning

Discrete least-squares approximations over optimized downward closed polynomial spaces in arbitrary dimension

Classical Force Fields Tailored for QM Applications: Is It Really a Feasible Strategy?

Learning-based subsampling

Sampling and Exact Reconstruction of Pulses with Variable Width

Just Sort It! A Simple and Effective Approach to Active Preference Learning

Geometric Integrators for Nonadiabatic Molecular Quantum Dynamics Induced by the Interaction with the Electromagnetic Field

Just Sort It! A Simple and Effective Approach to Active Preference Learning

On the performance of random reshuffling in stochastic learning

Discrete least-squares approximations over optimized downward closed polynomial spaces in arbitrary dimension

Classical Force Fields Tailored for QM Applications: Is It Really a Feasible Strategy?

Learning-based subsampling

Safe Adaptive Importance Sampling

Sampling at unknown locations, with an application in surface retrieval

Sampling and Exact Reconstruction of Pulses with Variable Width

Genetic Optimization of Training Sets for Improved Machine Learning Models of Molecular Properties