arXiv subjects

Jan Hamaekers

Publications and source records attributed to Jan Hamaekers.

15 recordsLinked to original sources

arXiv2026

Sparse Configuration Interaction for the Electronic Schr\"odinger Equation Revisited: Complete Basis Set Limit Complexity and Quantum-Encoding Impact

In this article we revisit regularity results for eigenfunctions in the discrete spectrum of the electronic Schr\"odinger equation and study their consequences for approximation complexity. In particular, for the convergence to the complete basis set limit, it can be shown that the curse of dimensionality in the leading algebraic exponent can be mitigated. That is, for general sparse grid constructions, the main term of the convergence rate with respect to the number of degrees of freedom is independent of the number of electrons. These insights indicate potential benefits for classical numerical solvers of the electronic Schr\"odinger equation and also for quantum-computing approaches through new qubit-efficient wavefunction encodings.

quant-ph↗

arXiv2026

Gaussian Process Regression-based Knowledge Distillation Framework for Simultaneous Prediction of Physical and Mechanical Properties of Epoxy Polymers

Epoxy polymers are widely used due to their multifunctional properties, but machine learning (ML) applications remain limited owing to their complex 3D molecular structure, multi-component nature, and lack of curated datasets. Existing ML studies are largely restricted to simulation data, specific properties, or narrow constituent ranges. To address these limitations, we developed an informed Gaussian Process Regression-based Knowledge Distillation (GPR-KD) framework for predicting multiple physical (glass transition temperature, density) and mechanical properties (elastic modulus, tensile strength, compressive strength, flexural strength, fracture energy, adhesive strength) of thermoset epoxy polymers. The model was trained on experimental literature data covering diverse monomer classes (9 resins, 40 hardeners). Individual GPR models serve as teacher models capturing nonlinear feature-property relationships, while a unified neural network student model learns distilled knowledge across all properties simultaneously. By encoding the target property as an input feature, the student model leverages cross-property correlations. Molecular-level descriptors extracted from SMILES representations using RDKit create a physics-informed model. The framework combines GPR interpretability and robustness with deep learning scalability and generalization. Comparative analysis demonstrates superior prediction accuracy over conventional ML models. Simultaneous multi-property prediction further improves accuracy through information sharing across correlated properties. The proposed framework enables accelerated design of novel epoxy polymers with tailored properties.

cond-mat.soft↗

arXiv2026

Flexible Cutoff Learning: Optimizing Machine Learning Potentials After Training

We introduce Flexible Cutoff Learning (FCL), a method for training machine learning interatomic potentials (MLIPs) whose cutoff radii can be adjusted after training. Unlike conventional MLIPs that fix the cutoff radius during training, FCL models are trained by randomly sampling cutoff radii independently for each atom. The resulting model can then be deployed with different per-atom cutoff radii depending on the application, enabling application-specific optimization of the accuracy-cost tradeoff. Using a differentiable cost model, these per-atom cutoffs can be optimized for specific target systems after training. We demonstrate FCL with a modified MACE architecture trained on the MAD dataset. For a subset featuring molecular crystals, optimized per-atom cutoffs reduce computational cost by more than 60% while increasing force errors by less than 1%. These results show that FCL enables training of a single general-purpose MLIP that can be adapted to diverse applications through post-training cutoff optimization, eliminating the need for retraining.

cond-mat.mtrl-sci↗

arXiv2025

Materium: An Autoregressive Approach for Material Generation

We present Materium: an autoregressive transformer for generating crystal structures that converts 3D material representations into token sequences. These sequences include elements with oxidation states, fractional coordinates and lattice parameters. Unlike diffusion approaches, which refine atomic positions iteratively through many denoising steps, Materium places atoms at precise fractional coordinates, enabling fast, scalable generation. With this design, the model can be trained in a few hours on a single GPU and generate samples much faster on GPUs and CPUs than diffusion-based approaches. The model was trained and evaluated using multiple properties as conditions, including fundamental properties, such as density and space group, as well as more practical targets, such as band gap and magnetic density. In both single and combined conditions, the model performs consistently well, producing candidates that align with the requested inputs.

cs.LG↗

arXiv2024

On Multilevel Energy-Based Fragmentation Methods

Energy-based fragmentation methods approximate the potential energy of a molecular system as a sum of contribution terms built from the energies of particular subsystems. Some such methods reduce to truncations of the many-body expansion (MBE); others combine subsystem energies in a manner inspired by the principle of inclusion/exclusion (PIE). The combinatorial technique of M\"obius inversion of sums over partially ordered sets, which generalizes the PIE, is known to provide a non-recursive expression for the MBE contribution terms, and has also been connected to related cluster expansion methods. We build from these ideas a very general framework for decomposing potential functions into energetic contribution terms associated with elements of particular partially ordered sets (posets) and direct products thereof. Specific choices immediately reproduce not only the MBE, but also a number of other existing decomposition forms, including, e.g., the multilevel ML-BOSSANOVA schema. Furthermore, a different choice of poset product leads to a setup familiar from the combination technique for high-dimensional approximation, which has a known connection to quantum-chemical composite methods. We present the ML-SUPANOVA decomposition form, which allows the further refinement of the terms of an MBE-like expansion of the Born-Oppenheimer potential according to systematic hierarchies of ab initio methods and of basis sets. We outline an adaptive algorithm for the a posteori construction of quasi-optimal truncations of this decomposition. Some initial experiments are reported and discussed.

math.NA↗

arXiv2023

Feature-based prediction of properties of cross-linked epoxy polymers by molecular dynamics and machine learning techniques

Epoxy polymers are used in wide range of applications. The properties and performance of epoxy polymers depend upon various factors like the type of constituents and their proportions used and other process parameters. The conventional way of developing epoxy polymers is usually labor-intensive and may not be fully efficient, which has resulted in epoxy polymers having a limited performance range due to the use of predetermined blend combinations, compositions and development parameters. Hence, in order to experiment with more design parameters, robust and easy computational techniques need to be established. To this end, we developed and analyzed in this study a new machine learning (ML) based approach to predict the mechanical properties of epoxy polymers based on their basic structural features. The results from molecular dynamics (MD) simulations have been used to derive the ML model. The salient feature of our work is that for the development of epoxy polymers based on EPON-862, several new hardeners were explored in addition to the conventionally used ones. The influence of additional parameters like the proportion of curing agent used and the extent of curing on the mechanical properties of epoxy polymers were also investigated. This method can be further extended by providing the epoxy polymer with the desired properties through knowledge of the structural characteristics of its constituents. The findings of our study can thus lead toward development of efficient design methodologies for epoxy polymeric systems.

cond-mat.mtrl-sci↗

arXiv2023

LLamol: A Dynamic Multi-Conditional Generative Transformer for De Novo Molecular Design

Generative models have demonstrated substantial promise in Natural Language Processing (NLP) and have found application in designing molecules, as seen in General Pretrained Transformer (GPT) models. In our efforts to develop such a tool for exploring the organic chemical space in search of potentially electro-active compounds, we present "LLamol", a single novel generative transformer model based on the LLama 2 architecture, which was trained on a 13M superset of organic compounds drawn from diverse public sources. To allow for a maximum flexibility in usage and robustness in view of potentially incomplete data, we introduce "Stochastic Context Learning" as a new training procedure. We demonstrate that the resulting model adeptly handles single- and multi-conditional organic molecule generation with up to four conditions, yet more are possible. The model generates valid molecular structures in SMILES notation while flexibly incorporating three numerical and/or one token sequence into the generative process, just as requested. The generated compounds are very satisfactory in all scenarios tested. In detail, we showcase the model's capability to utilize token sequences for conditioning, either individually or in combination with numerical properties, making LLamol a potent tool for de novo molecule design, easily expandable with new properties.

cs.LG↗

arXiv2023

Predicting Properties of Oxide Glasses Using Informed Neural Networks

Many modern-day applications require the development of new materials with specific properties. In particular, the design of new glass compositions is of great industrial interest. Current machine learning methods for learning the composition-property relationship of glasses promise to save on expensive trial-and-error approaches. Even though quite large datasets on the composition of glasses and their properties already exist (i.e., with more than 350,000 samples), they cover only a very small fraction of the space of all possible glass compositions. This limits the applicability of purely data-driven models for property prediction purposes and necessitates the development of models with high extrapolation power. In this paper, we propose a neural network model which incorporates prior scientific and expert knowledge in its learning pipeline. This informed learning approach leads to an improved extrapolation power compared to blind (uninformed) neural network models. To demonstrate this, we train our models to predict three different material properties, that is, the glass transition temperature, the Young's modulus (at room temperature), and the shear modulus of binary oxide glasses which do not contain sodium. As representatives for conventional blind neural network approaches we use five different feed-forward neural networks of varying widths and depths. For each property, we set up model ensembles of multiple trained models and show that, on average, our proposed informed model performs better in extrapolating the three properties of previously unseen sodium borate glass samples than all five conventional blind models.

physics.comp-ph↗

arXiv2023

On the Interplay of Subset Selection and Informed Graph Neural Networks

Machine learning techniques paired with the availability of massive datasets dramatically enhance our ability to explore the chemical compound space by providing fast and accurate predictions of molecular properties. However, learning on large datasets is strongly limited by the availability of computational resources and can be infeasible in some scenarios. Moreover, the instances in the datasets may not yet be labelled and generating the labels can be costly, as in the case of quantum chemistry computations. Thus, there is a need to select small training subsets from large pools of unlabelled data points and to develop reliable ML methods that can effectively learn from small training sets. This work focuses on predicting the molecules atomization energy in the QM9 dataset. We investigate the advantages of employing domain knowledge-based data sampling methods for an efficient training set selection combined with informed ML techniques. In particular, we show how maximizing molecular diversity in the training set selection process increases the robustness of linear and nonlinear regression techniques such as kernel methods and graph neural networks. We also check the reliability of the predictions made by the graph neural network with a model-agnostic explainer based on the rate distortion explanation framework.

physics.chem-ph↗

arXiv2023

Parameterized Neural Networks for Finance

We discuss and analyze a neural network architecture, that enables learning a model class for a set of different data samples rather than just learning a single model for a specific data sample. In this sense, it may help to reduce the overfitting problem, since, after learning the model class over a larger data sample consisting of such different data sets, just a few parameters need to be adjusted for modeling a new, specific problem. After analyzing the method theoretically and by regression examples for different one-dimensional problems, we finally apply the approach to one of the standard problems asset managers and banks are facing: the calibration of spread curves. The presented results clearly show the potential that lies within this method. Furthermore, this application is of particular interest to financial practitioners, since nearly all asset managers and banks which are having solutions in place may need to adapt or even change their current methodologies when ESG ratings additionally affect the bond spreads.

q-fin.ST↗

arXiv2021

Similarity of particle systems using an invariant root mean square deviation measure

Determining whether two particle systems are similar is a common problem in particle simulations. When the comparison should be invariant under permutations, orthogonal transformations, and translations of the systems, special techniques are needed. We present an algorithm that can test particle systems of finite size for similarity and, if they are similar, can find the optimal alignment between them. Our approach is based on an invariant version of the root mean square deviation (RMSD) measure and is capable of finding the globally optimal solution in $O(n^3)$ operations where $n$ is the number of three-dimensional particles.

cond-mat.stat-mech↗

arXiv2017

The octet rule in chemical space: Generating virtual molecules

We present a generator of virtual molecules that selects valid chemistry on the basis of the octet rule. Also, we introduce a mesomer group key that allows a fast detection of duplicates in the generated structures. Compared to existing approaches, our model is simpler and faster, generates new chemistry and avoids invalid chemistry. Its versatility is illustrated by the correct generation of molecules containing third-row elements and a surprisingly adept handling of complex boron chemistry. Without any empirical parameters, our model is designed to be valid also in unexplored regions of chemical space. One first unexpected finding is the high prevalence of dipolar structures among generated molecules.

physics.chem-ph↗

arXiv2017

ATK-ForceField: A New Generation Molecular Dynamics Software Package

ATK-ForceField is a software package for atomistic simulations using classical interatomic potentials. It is implemented as a part of the Atomistix ToolKit (ATK), which is a Python programming environment that makes it easy to create and analyze both standard and highly customized simulations. This paper will focus on the atomic interaction potentials, molecular dynamics, and geometry optimization features of the software, however, many more advanced modeling features are available. The implementation details of these algorithms and their computational performance will be shown. We present three illustrative examples of the types of calculations that are possible with ATK-ForceField: modeling thermal transport properties in a silicon germanium crystal, vapor deposition of selenium molecules on a selenium surface, and a simulation of creep in a copper polycrystal.

cond-mat.mtrl-sci↗

arXiv2016

Localized Coulomb Descriptors for the Gaussian Approximation Potential

We introduce a novel class of localized atomic environment representations, based upon the Coulomb matrix. By combining these functions with the Gaussian approximation potential approach, we present LC-GAP, a new system for generating atomic potentials through machine learning (ML). Tests on the QM7, QM7b and GDB9 biomolecular datasets demonstrate that potentials created with LC-GAP can successfully predict atomization energies for molecules larger than those used for training to chemical accuracy, and can (in the case of QM7b) also be used to predict a range of other atomic properties with accuracy in line with the recent literature. As the best-performing representation has only linear dimensionality in the number of atoms in a local atomic environment, this represents an improvement both in prediction accuracy and computational cost when considered against similar Coulomb matrix-based methods.

stat.ML↗

arXiv2016

An Adaptive Multiscale Approach for Electronic Structure Methods

In this paper, we introduce a new scheme for the efficient numerical treatment of the electronic Schr\"odinger equation for molecules. It is based on the combination of a many-body expansion, which corresponds to the so-called bond order dissection Anova approach, with a hierarchy of basis sets of increasing order. Here, the energy is represented as a finite sum of contributions associated to subsets of nuclei and basis sets in a telescoping sum like fashion. Under the assumption of data locality of the electronic density (nearsightedness of electronic matter), the terms of this expansion decay rapidly and higher terms may be neglected. We further extend the approach in a dimension-adaptive fashion to generate quasi-optimal approximations, i.e. a specific truncation of the hierarchical series such that the total benefit is maximized for a fixed amount of costs. This way, we are able to achieve substantial speed up factors compared to conventional first principles methods depending on the molecular system under consideration. In particular, the method can deal efficiently with molecular systems which include only a small active part that needs to be described by accurate but expensive models.

math.NA↗