Search

Zero and ultralow-field nuclear magnetic resonance (ZULF NMR) is an NMR modality where experiments are performed in fields at which spin-spin interactions within molecules and materials are stronger than Zeeman interactions. This typically occurs at external fields of microtesla strength or below, considerably smaller than Earth's field. In ZULF NMR, the measurement of spin-spin couplings and spin relaxation rates provides a nondestructive means for identifying chemicals and chemical fragments, and for conducting sample or process analyses. The absence of the symmetry imposed by a strong external magnetic field enables experiments that exploit terms in the nuclear spin Hamiltonian that are suppressed in high-field NMR, which in turn opens up new capabilities in a broad range of fields, from the search for dark matter to the preparation of hyperpolarized contrast agents for clinical imaging. Furthermore, as in ZULF NMR the Larmor frequencies are typically in the audio band, the nuclear spins can be addressed with d.c. magnetic field pulses, and highly sensitive magnetometers are used for detection. In contrast to high-field NMR, the low-frequency signals readily pass through conductive materials such as metals, and heterogeneous samples do not produce resonance line broadening, meaning that high-resolution spectroscopy is possible. Notable practical advantages of ZULF NMR spectroscopy are the low cost and relative simplicity and portability of the spectrometer system. In recent years ZULF NMR has become more accessible, thanks to improvements in magnetometer sensitivity and their commercial availability, and the development of hyperpolarization methods that provide a simple means to boost signal strengths by several orders of magnitude. These topics are reviewed and a perspective on potential future avenues of ZULF-NMR research is presented., Comment: 71 page, 44 figures, 425 references

Numerical weather prediction (NWP) models often underperform compared to simpler climatology-based precipitation forecasts in northern tropical Africa, even after statistical postprocessing. AI-based forecasting models show promise but have avoided precipitation due to its complexity. Synoptic-scale forcings like African easterly waves and other tropical waves (TWs) are important for predictability in tropical Africa, yet their value for predicting daily rainfall remains unexplored. This study uses two machine-learning models--gamma regression and a convolutional neural network (CNN)--trained on TW predictors from satellite-based GPM IMERG data to predict daily rainfall during the July-September monsoon season. Predictor variables are derived from the local amplitude and phase information of seven TW from the target and up-and-downstream neighboring grids at 1-degree spatial resolution. The ML models are combined with Easy Uncertainty Quantification (EasyUQ) to generate calibrated probabilistic forecasts and are compared with three benchmarks: Extended Probabilistic Climatology (EPC15), ECMWF operational ensemble forecast (ENS), and a probabilistic forecast from the ENS control member using EasyUQ (CTRL EasyUQ). The study finds that downstream predictor variables offer the highest predictability, with downstream tropical depression (TD)-type wave-based predictors being most important. Other waves like mixed-Rossby gravity (MRG), Kelvin, and inertio-gravity waves also contribute significantly but show regional preferences. ENS forecasts exhibit poor skill due to miscalibration. CTRL EasyUQ shows improvement over ENS and marginal enhancement over EPC15. Both gamma regression and CNN forecasts significantly outperform benchmarks in tropical Africa. This study highlights the potential of ML models trained on TW-based predictors to improve daily precipitation forecasts in tropical Africa.

Optimization problems with random data have a wide range of applications. A typical feature of many such problems is that some variables have to be optimized before certain random coefficients have been realized and for other variables it is sufficient to decide on them afterwards. This leads to a multiple stage decision process. To optimize the variables in the first of two subsequent stages the stochastic problem is transformed into a deterministic program, called the (deterministic equivalent of the) recourse problem. In case of stochastic linear programs with finitely distributed random data this non-stochastic substitute is just a linear program. In the same way a multiple objective linear program is obtained if the original problem has multiple objective functions. In the first of the two stages, a decision maker usually would chose a feasible point out of the set of all Pareto-optimal points. This choice however has consequences to later stage decisions. We claim that the decision process in the earlier of the two stages is not fully transparent if a classical multi-objective decision process is applied: in addition to the original objectives of the problem a decision maker may have a preference for largest possible flexibility in later stage decisions. This additional objective is taken into account if the recourse problem in case of multiple objectives is taken to be a polyhedral convex set optimization problem instead of a multi-objective linear program only. We also discuss several surrogate problems to the recourse problem such as the wait-and-see problem and the expected valued problem for the multi-objective case. The new approach based on set optimization is illustrated by an example, the multi-objective newsvendor problem.

Numerical weather prediction (NWP) models struggle to skillfully predict tropical precipitation occurrence and amount, calling for alternative approaches. For instance, it has been shown that fairly simple, purely data-driven logistic regression models for 24-hour precipitation occurrence outperform both climatological and NWP forecasts for the West African summer monsoon. More complex neural network based approaches, however, remain underdeveloped due to the non-Gaussian character of precipitation. In this study, we develop, apply and evaluate a novel two-stage approach, where we train a U-Net convolutional neural network (CNN) model on gridded rainfall data to obtain a deterministic forecast and then apply the recently developed, nonparametric Easy Uncertainty Quantification (EasyUQ) approach to convert it into a probabilistic forecast. We evaluate CNN+EasyUQ for one-day ahead 24-hour accumulated precipitation forecasts over northern tropical Africa for 2011--2019, with the Integrated Multi-satellitE Retrievals for GPM (IMERG) data serving as ground truth. In the most comprehensive assessment to date we compare CNN+EasyUQ to state-of-the-art physics-based and data-driven approaches such as a monthly probabilistic climatology, raw and postprocessed ensemble forecasts from the European Centre for Medium-Range Weather Forecasts (ECMWF), and traditional statistical approaches that use up to 25 predictor variables from IMERG and the ERA5 reanalysis.Generally, statistical approaches perform about en par with post-processed ECMWF ensemble forecasts. The CNN+EasyUQ approach, however, clearly outperforms all competitors for both occurrence and amount. Hybrid methods that merge CNN+EasyUQ and physics-based forecasts show slight further improvement. Thus, the CNN+EasyUQ approach can likely improve operational probabilistic forecasts of rainfall in the tropics, and potentially even beyond.

Motivated by situations in which the removal of a zero (a.k.a., an absorbing element) from a semigroup yields a subsemigroup with another zero, sets of quasi-zeros (a.k.a., quasi-absorbing elements) are introduced as well as primitive elements, minimal ideals and simple semigroups all with respect to a given ideal. An application to set optimization is discussed., Comment: 17 pages

Recently defined expectile regions capture the idea of centrality with respect to a multivariate distribution, but fail to describe the tail behavior while it is not at all clear what should be understood by a tail of a multivariate distribution. Therefore, cone expectile sets are introduced which take into account a vector preorder for the multi-dimensional data points. This provides a way of describing and clustering a multivariate distribution/data cloud with respect to an order relation. Fundamental properties of cone expectiles including dual representations of both expectile regions and cone expectile sets are established. It is shown that set-valued sublinear risk measures can be constructed from cone expectile sets in the same way as in the univariate case. Inverse functions of cone expectiles are defined which should be considered as rank functions rather than depth functions. Finally, expectile orders for random vectors are introduced and characterized via expectile rank functions.

Cone distribution functions from statistics are turned into Multi-Criteria Decision Making tools. It is demonstrated that this procedure can be considered as an upgrade of the weighted sum scalarization insofar as it absorbs a whole collection of weighted sum scalarizations at once instead of fixing a particular one in advance. As examples show, this type of scalarization--in contrast to a pure weighted sum scalarization-is also able to detect ``non-convex" parts of the Pareto frontier. Situations are characterized in which different types of rank reversal occur, and it is explained why this might even be useful for analyzing the ranking procedure. The ranking functions are then extended to sets providing unary indicators for set preferences which establishes, for the first time, the link between set optimization methods and set-based multi-objective optimization. A potential application in machine learning is outlined., Comment: 20 pages, 12 figures

In this paper, we describe the capabilities and constraints of Large Language Models (LLMs) within disparate academic disciplines, aiming to delineate their strengths and limitations with precision. We examine how LLMs augment scientific inquiry, offering concrete examples such as accelerating literature review by summarizing vast numbers of publications, enhancing code development through automated syntax correction, and refining the scientific writing process. Simultaneously, we articulate the challenges LLMs face, including their reliance on extensive and sometimes biased datasets, and the potential ethical dilemmas stemming from their use. Our critical discussion extends to the varying impacts of LLMs across fields, from the natural sciences, where they help model complex biological sequences, to the social sciences, where they can parse large-scale qualitative data. We conclude by offering a nuanced perspective on how LLMs can be both a boon and a boundary to scientific progress.

An important aspect in the development of small molecules as drugs or agro-chemicals is their systemic availability after intravenous and oral administration. The prediction of the systemic availability from the chemical structure of a potential candidate is highly desirable, as it allows to focus the drug or agrochemical development on compounds with a favorable kinetic profile. However, such pre-dictions are challenging as the availability is the result of the complex interplay between molecular properties, biology and physiology and training data is rare. In this work we improve the hybrid model developed earlier [1]. We reduce the median fold change error for the total oral exposure from 2.85 to 2.35 and for intravenous administration from 1.95 to 1.62. This is achieved by training on a larger data set, improving the neural network architecture as well as the parametrization of mechanistic model. Further, we extend our approach to predict additional endpoints and to handle different covariates, like sex and dosage form. In contrast to a pure machine learning model, our model is able to predict new end points on which it has not been trained. We demonstrate this feature by predicting the exposure over the first 24h, while the model has only been trained on the total exposure., Comment: Version accepted by Journal of Computer-Aided Molecular Design