25 Papers in Highly-Ranked Journals

High-performance batteries need accelerated discovery and optimization of new anode materials. Herein, we explore the Si─Ge─Sn ternary alloy system as a candidate fast-charging anode materials system by utilizing a scanning droplet cell (SDC) as an autonomous electrochemical characterization tool with the goal of subsequent upscaling. As the SDC is performing experiments sequentially, an exploration of the entire ternary space is unfeasible due to time constraints. Thus, closed-loop optimization, guided by real-time data analysis and sequential learning algorithms, is utilized to direct experiments. The lead material identified is scaled up to a coin cell to validate the findings from the autonomous millimeter-scale thin-film electrochemical experimentation. Explainable machine learning (ML) models incorporating data from high-throughput Raman spectroscopy and X-ray diffraction (XRD) are used to elucidate the effect of short and long-range ordering on material performance.

Chemical ionization mass spectrometry (CIMS) is widely used in atmospheric chemistry studies. However, due to the complex interactions between reagent ions and target compounds, chemical understanding remains limited and compound identification difficult. In this study, we apply machine learning to a reference dataset of pesticides in two standard solutions to build a model that can provide insights from CIMS analyses in atmospheric science. The CIMS measurements were performed with an Orbitrap mass spectrometer coupled to a thermal desorption multi-scheme chemical ionization inlet unit (TD-MION-MS) with both negative and positive ionization modes utilizing Br−, , H3O+ and (CH3)2COH+ (AceH+) as reagent ions. We then trained two machine learning methods on these data: (1) random forest (RF) for classifying if a pesticide can be detected with CIMS and (2) kernel ridge regression (KRR) for predicting the expected CIMS signals. We compared their performance on five different representations of the molecular structure: the topological fingerprint (TopFP), the molecular access system keys (MACCS), a custom descriptor based on standard molecular properties (RDKitPROP), the Coulomb matrix (CM) and the many-body tensor representation (MBTR). The results indicate that MACCS outperforms the other descriptors. Our best classification model reaches a prediction accuracy of 0.85 ± 0.02 and a receiver operating characteristic curve area of 0.91 ± 0.01. Our best regression model reaches an accuracy of 0.44 ± 0.03 logarithmic units of the signal intensity. Subsequent feature importance analysis of the classifiers reveals that the most important sub-structures are NH and OH for the negative ionization schemes and nitrogen-containing groups for the positive ionization schemes.

Under the mounting pressure from global warming, green roofs emerge as a valuable source for climate adaptation, particularly in compact metropolises where green space is limited. Consequently, there is a need to quantitatively evaluate the potential for roof greening where it is most needed and suitable. Despite the increasing importance of this issue, there have been limited studies on the effectiveness of remote sensing and deep learning in identifying the potential for roof greening in many cities. To address this, we have created a GreenRoof dataset, comprising approximately 6400 pairs of remote sensing images and corresponding masks of roofs with high greening potential in four European cities. Afterward, we exploit the capabilities of deep learning methods to identify roofs that are suitable for greening from remote sensing images. Using 15 German cities as a case study for future urban rooftop planning, we estimate the spatial potential for retrofitting green roofs. Structural parameters for prioritizing green roof implementation include vegetation coverage, thermal environment, and building density. Results indicate that the total area suitable for green roof retrofitting exceeds 20% of the roof area in the 15 German cities examined. The spatial analysis effectively reflects variation in demand and suitability for green roof retrofitting across different cities. In conclusion, this study provides a versatile screening approach utilizing remote sensing, deep learning, and spatial analysis, which can be readily adapted to inform municipal policies in other cities aiming to promote green roofs and enhance sustainable urban development.

Crime is responsible for major financial losses and serious harm to the well-being of individuals, and, hence, a crucial task of police operations is effective patrolling. Yet, in existing decision models aimed at police operations, microscopic routing decisions from patrolling are not considered, and, furthermore, the objective is limited to surrogate metrics (e. g., response time) instead of crime prevention. In this paper, we thus formalize the decision problem of dynamic police patrolling as a Markov decision process that models microscopic routing decisions, so that the expected number of prevented crimes are maximized. We experimentally show that standard solution approaches for our decision problem are not scalable to real-world settings. As a remedy, we present a tailored and highly efficient Monte Carlo tree search algorithm. We then demonstrate our algorithm numerically using real-world crime data from Chicago and show that the decision-making by our algorithm offers significant improvements for crime prevention over patrolling tactics from current practice. Informed by our results, we finally discuss implications for improving the patrolling tactics in police operations.

This study investigates the predictive capability of radiomics in determining programmed cell death ligand 1 (PD-L1) expression (>=1%) status in non-small cell lung cancer (NSCLC) patients using a newly collected [18F]FDG PET/CT dataset. We aimed to replicate and validate the radiomics-based machine learning (ML) model proposed by Zhao et al. [2] predicting PD-L1 status from PET/CT-imaging.
An independent cohort of 254 NSCLC patients underwent [18F]FDG PET/CT imaging, with primary tumor segmentation conducted using lung tissue window (LTW) and more conservative soft tissue window (STW) methods. Radiomics models (“Rad-score” and “complex model”) and a clinical-stage model from Zhao et al. were evaluated via 10-fold cross-validation and AUC analysis, alongside a benchmark-study comparing different ML-model pipelines. Clinicopathological data were collected from medical records.
On our data, the Rad-score model yielded mean AUCs of 0.593 (STW) and 0.573 (LTW), below Zhao et al.’s 0.761. The complex model achieved mean AUCs of 0.505 (STW) and 0.519 (LTW), lower than Zhao et al.’s 0.769. The clinical model showed a mean AUC of 0.555, below Zhao et al.’s 0.64. All models performed significantly lower than Zhao et al.’s findings. Our benchmark study on four ML pipelines revealed consistently low performance across all configurations.
Our study failed to replicate original findings, suggesting poor model performance and questioning predictive value of radiomics features in classifying PD-L1 expression from PET/CT imaging. These results highlight challenges in replicating radiomics-based ML models and stress the need for rigorous validation

Objectives: Adenomatous colorectal polyps require endoscopic resection, as opposed to non-adenomatous hyperplastic colorectal polyps. This study aims to evaluate the effect of artificial intelligence (AI)-assisted differentiation of adenomatous and non-adenomatous colorectal polyps at CT colonography on radiologists’ therapy management.
Materials and methods: Five board-certified radiologists evaluated CT colonography images with colorectal polyps of all sizes and morphologies retrospectively and decided whether the depicted polyps required endoscopic resection. After a primary unassisted reading based on current guidelines, a second reading with access to the classification of a radiomics-based random-forest AI-model labelling each polyp as ’non-adenomatous’ or ‘adenomatous’ was performed. Performance was evaluated using polyp histopathology as the reference standard.
Results: 77 polyps in 59 patients comprising 118 polyp image series (47% supine position, 53% prone position) were evaluated unassisted and AI-assisted by five independent board-certified radiologists, resulting in a total of 1180 readings (subsequent polypectomy: yes or no). AI-assisted readings had higher accuracy (76% +/− 1% vs. 84% +/− 1%), sensitivity (78% +/− 6% vs. 85% +/− 1%), and specificity (73% +/− 8% vs. 82% +/− 2%) in selecting polyps eligible for polypectomy (p < 0.001). Inter-reader agreement was improved in the AI-assisted readings (Fleiss’ kappa 0.69 vs. 0.92).
Conclusion: AI-based characterisation of colorectal polyps at CT colonography as a second reader might enable a more precise selection of polyps eligible for subsequent endoscopic resection. However, further studies are needed to confirm this finding and histopathologic polyp evaluation is still mandatory.

The development of affective computing and medical electronic technologies has led to the emergence of Artificial Intelligence (AI)-based methods for the early detection of depression. However, previous studies have often overlooked the necessity for the AI-assisted diagnosis system to be wearable and accessible in practical scenarios for depression recognition. In this work, we present an on-board executable multi-feature transfer-enhanced fusion model for our custom-designed wearable three-lead Electroencephalogram (EEG) sensor, based on EEG data collected from 73 depressed patients and 108 healthy controls. Experimental results show that the proposed model exhibits low-computational complexity (65.0 K parameters), promising Floating-Point Operations (FLOPs) performance (25.6 M), real-time processing (1.5 s/execution), and low power consumption (320.8 mW). Furthermore, it requires only 202.0 KB of Random Access Memory (RAM) and 279.6 KB of Read-Only Memory (ROM) when deployed on the EEG sensor. Despite its low computational and spatial complexity, the model achieves a notable classification accuracy of 95.2%, specificity of 94.0%, and sensitivity of 96.9% under independent test conditions. These results underscore the potential of deploying the model on the wearable three-lead EEG sensor for assisting in the diagnosis of depression.

High-quality annotations are a critical success factor for machine learning (ML) applications. To achieve this, we have
traditionally relied on human annotators, navigating the challenges of limited budgets and the varying task-specific expertise, costs, and availability. Since the emergence of Large Language Models (LLMs), their popularity for generating automated annotations has grown, extending possibilities and complexity of designing an efficient annotation strategy. Increasingly, computer vision capabilities have been integrated into general-purpose LLMs like ChatGPT. This raises the question of how effectively LLMs can be used in satellite image annotation tasks and how they compare to traditional annotator types. This study presents a comprehensive investigation and comparison of various human and automated annotators for image classification. We evaluate the feasibility and economic competitiveness of using the ChatGPT4-V model for a complex land usage annotation task and compare it with alternative human annotators. A set of satellite images is annotated by a domain expert and 15 additional human and automated annotators, differing in expertise and costs. Our analyses examine the annotation quality loss between the expert and other annotators. This comparison is conducted through (1) descriptive analyses, (2) fitting linear probability models, and (3) comparing F1-scores. Ultimately, we simulate annotation strategies where samples are split according to an automatically assigned certainty score. Routing low-certainty images to human annotators can cut total annotation costs by over 50% with minimal impact on label quality. We discuss implications regarding the economic competitiveness of annotation strategies, prompt engineering and the task-specificity of expertise.

Medical ultrasound has been widely used to examine vascular structure in modern clinical practice. However, traditional ultrasound examination often faces challenges related to inter- and intra-operator variation. The robotic ultrasound system (RUSS) appears as a potential solution for such challenges because of its superiority in stability and reproducibility. Given the complex anatomy of human vasculature, multiple vessels often appear in ultrasound images, or a single vessel bifurcates into branches, complicating the examination process. To tackle this challenge, this work presents a gaze-guided RUSS for vascular applications. A gaze tracker captures the eye movements of the operator. The extracted gaze signal guides the RUSS to follow the correct vessel when it bifurcates. Additionally, a gaze-guided segmentation network is proposed to enhance segmentation robustness by exploiting gaze information. However, gaze signals are often noisy, requiring interpretation to accurately discern the operator’s true intentions. To this end, this study proposes a stabilization module to process raw gaze data. The inferred attention heatmap is utilized as a region proposal to aid segmentation and serve as a trigger signal when the operator needs to adjust the scanning target, such as when a bifurcation appears. To ensure appropriate contact between the probe and surface during scanning, an automatic ultrasound confidence-based orientation correction method is developed. In experiments, we demonstrated the efficiency of the proposed gaze-guided segmentation pipeline by comparing it with other methods. Besides, the performance of the proposed gaze-guided RUSS was also validated as a whole on a realistic arm phantom with an uneven surface.

Foundation models are widely utilised for their strong representational capabilities, driven by training on extensive datasets with self-supervised learning. The increasing complexity of these models highlights the importance of interpretability to enhance transparency and improve human understanding of their decision-making processes. Most existing interpretability methods explain model behaviour by attributing importance to individual data elements across different layers, based on their influence on the final prediction. These approaches often emphasise only the most relevant features, overlooking the broader representational space, removing less important features. In this study, we propose a novel framework for explanation generation that serves as an alternative to feature removal, offering a more comprehensive understanding of model behaviour. Our framework leverages the generative abilities of audio language models to replace removed features with contextually appropriate alternatives, providing a more complete view of the model’s decision-making process. Through extensive evaluations on standard benchmarks, including keyword spotting and speech emotion recognition, our approach demonstrates its effectiveness in generating high-quality audio explanations.

Automated Depression Detection (ADD) in speech aims to automatically estimate one’s depressive attributes through artificial intelligence tools towards spoken signals. Nevertheless, existing speech-based ADD works fail to sufficiently consider weakly-supervised cases with inaccurate labels, which may typically appear in intelligent mental health. In this regard, we propose the Self-Learning-based Label Correction (SLLC) approach for weakly-supervised depression detection in speech. The proposed approach employs a self-learning manner connecting a label correction module and a depression detection module. Within the approach, the label correction module fuses likelihood-ratio-based and prototype-based label correction strategies in order to effectively correct the inaccurate labels, while the depression detection module aims at detecting depressed samples through a 1D convolutional recurrent neural network with multiple types of losses. The experimental results on two depression detection corpora show that our proposed SLLC approach performs better compared with existing state-of-the-art speech-based depression detection approaches, in the case of weak supervision with inaccurate labels for depression detection in speech.

Accurate extraction of building footprints from satellite imagery is of high value. Currently, deep learning methods are predominant in this field due to their powerful representation capabilities. However, they generally require extensive pixel-wise annotations, which constrains their practical application. Semi-supervised learning (SSL) significantly mitigates this requirement by leveraging large volumes of unlabeled data for model self-training (ST), thus enhancing the viability of building footprint extraction. Despite its advantages, SSL faces a critical challenge: the imbalanced distribution between the majority background class and the minority building class, which often results in model bias toward the background during training. To address this issue, this article introduces a novel method called DeBiased matching (DBMatch) for semi-supervised building footprint extraction. DBMatch comprises three main components: 1) a basic supervised learning module (SUP) that uses labeled data for initial model training; 2) a classical weak-to-strong ST module that generates pseudo-labels from unlabeled data for further model ST; and 3) a novel logit debiasing (LDB) module that calculates a global logit bias between building and background, allowing for dynamic pseudo-label calibration. To verify the effectiveness of the proposed DBMatch, extensive experiments are performed on three public building footprint extraction datasets covering six global cities in SSL setting. The experimental results demonstrate that our method significantly outperforms some advanced SSL methods in semi-supervised building footprint extraction.

Earth observation (EO) has inevitably entered the Big Data era. The computational challenge associated with analyzing large EO data using sophisticated deep learning models has become a significant bottleneck. To address this challenge, there has been a growing interest in exploring quantum computing as a potential solution. However, the process of encoding EO data into quantum states for analysis potentially undermines the efficiency advantages gained from quantum computing. This article introduces a hybrid quantum deep learning model that effectively encodes and analyzes EO data for classification tasks. The proposed model uses an efficient encoding approach called superpixel encoding, which reduces the quantum resources required for large image representation by incorporating the concept of superpixels. To validate the effectiveness of our model, we conducted evaluations on multiple EO benchmarks, including Overhead-MNIST, So2Sat LCZ42, and SAT-6 datasets. In addition, we studied the impacts of different interaction gates and measurements on classification performance to guide model optimization. The experimental results suggest the validity of our model for accurate classification of EO data.

Background: Chronic obstructive pulmonary disease (COPD) affects breathing, speech production, and coughing. We evaluated a machine learning analysis of speech for classifying the disease severity of COPD.
Methods: In this single centre study, non-consecutive COPD patients were prospectively recruited for comparing their speech characteristics during and after an acute COPD exacerbation. We extracted a set of spectral, prosodic, and temporal variability features, which were used as input to a support vector machine (SVM). Our baseline for predicting patient state was an SVM model using self-reported BORG and COPD Assessment Test (CAT) scores.
Results: In 50 COPD patients (52% males, 22% GOLD II, 44% GOLD III, 32% GOLD IV, all patients group E), speech analysis was superior in distinguishing during and after exacerbation status compared to BORG and CAT scores alone by achieving 84% accuracy in prediction. CAT scores correlated with reading rhythm, and BORG scales with stability in articulation. Pulmonary function testing (PFT) correlated with speech pause rate and speech rhythm variability.
Conclusion: Speech analysis may be a viable technology for classifying COPD status, opening up new opportunities for remote disease monitoring.

Age estimations are relevant for pre-trial detention, sentencing in criminal cases and as part of the evaluation in asylum processes to protect the rights and privileges of minors. No current method can determine an exact chronological age due to individual variations in biological development. This study seeks to develop a validated statistical model for estimating an age relative to key legal thresholds (15, 18, and 21 years) based on a skeletal (CT-clavicle, radiography-hand/wrist or MR-knee) and tooth (radiography-third molar) developmental stages. The whole model is based on 34 scientific studies, divided into examinations of the hand/wrist (15 studies), clavicle (5 studies), distal femur (4 studies), and third molars (10 studies). In total, data from approximately 27,000 individuals have been incorporated and the model has subsequently been validated with data from 5,000 individuals. The core framework of the model is built upon transition analysis and is further developed by a combination of a type of parametric bootstrapping and Bayesian theory. Validation of the model includes testing the models on independent datasets of individuals with known ages and shows a high precision with separate populations aligning closely with the model’s predictions. The practical use of the complex statistical model requires a user-friendly tool to provide probabilities together with the margin of error. The assessment based on the model forms the medical component for the overall evaluation of an individual’s age.

Land cover information is indispensable for advancing the United Nations’ sustainable development goals, and land cover mapping under a more detailed category system would significantly contribute to economic livelihood tracking and environmental degradation measurement. However, the substantial difficulty in acquiring fine-grained training data makes the implementation of this task particularly challenging. Here, we propose to combine fully labeled source domain and weakly labeled target domain for weakly supervised domain adaptation (WSDA). This is beneficial as the utilization of sparse and coarse weak labels can considerably alleviate the labor required for precise and detailed land cover annotation. Specifically, we introduce the Prototype-based pseudo-label Rectification and Expansion (PRE) approach, which leverages the prototypes (i.e., the class-wise feature centroids) as the bridge to connect sparse labels and global feature distributions. According to the feature distances to the prototypes, the confidence of pseudo-labels predicted in the unlabeled regions of the target domain is assessed. This confidence is then utilized to guide the dynamic expansion and rectification of pseudo-labels. Based on PRE, we carry out high categorical resolution land cover mapping for 10 cities in different regions around the world, severally using PlanetScope, Gaofen-1, and Sentinel-2 satellite images. In the study areas, we achieve cross-sensor, cross-category, and cross-continent WSDA, with the overall accuracy exceeding 80%. The promising results indicate that PRE is capable of reducing the dependency of land cover classification on high-quality annotations, thereby improving label efficiency. We expect our work to enable global fine-grained land cover mapping, which in turn promote Earth observation to provide more precise and thorough information for environmental monitoring.

I am grateful to James Cordeiro, Timothy Murphy, Heloise Robinson and Teresa Baron for their perceptive and stimulating comments on my article in this journal. In what follows, I seek to respond to some of the main points raised in each commentary.

Parenting our biological children is a centrally important matter, but how, if it all, can it be justified? According to a contemporary influential line of thinking, the acquisition by parents of a moral right to parent their biological children should be grounded by appeal to the value of the intimate emotional relationship that gestation facilitates between a newborn and a gestational procreator. I evaluate two arguments in defence of this proposal and argue that both are unconvincing.Data are available in a public, open access repository.

Cardiovascular diseases (CVD) can be diagnosed using various diagnostic modalities. The electrocardiogram (ECG) is a cost-effective and widely available diagnostic aid that provides functional information of the heart. However, its ability to classify and spatially localise CVD is limited. In contrast, cardiac magnetic resonance (CMR) imaging provides detailed structural information of the heart and thus enables evidence-based diagnosis of CVD, but long scan times and high costs limit its use in clinical routine. In this work, we present a deep learning strategy for cost-effective and comprehensive cardiac screening solely from ECG. Our approach combines multimodal contrastive learning with masked data modelling to transfer domain-specific information from CMR imaging to ECG representations. In extensive experiments using data from 40,044 UK Biobank subjects, we demonstrate the utility and generalisability of our method for subject-specific risk prediction of CVD and the prediction of cardiac phenotypes using only ECG data. Specifically, our novel multimodal pre-training paradigm improves performance by up to 12.19% for risk prediction and 27.59% for phenotype prediction. In a qualitative analysis, we demonstrate that our learned ECG representations incorporate information from CMR image regions of interest.

Despite significant progress in 3D medical image segmentation using deep learning, manual annotation remains a labor-intensive bottleneck. Self-supervised mask propagation (SMP) methods have emerged to alleviate this challenge, allowing intra-volume segmentation with just a single slice annotation. However, the previous SMP methods often rely on 2D information and ignore volumetric contexts. While our previous work, called Vol2Flow, attempts to address this concern, it exhibits limitations, including not focusing enough on local (i.e., slice-pair) information, neglecting global information (i.e., volumetric contexts) in the objective function, and error accumulation during slice-to-slice reconstruction. This paper introduces Flow2Mask, a novel SMP method, developed to overcome the limitations of previous SMP approaches, particularly Vol2Flow. During training, Flow2Mask proposes the Local-to-Global (L2G) loss to learn inter-slice flow fields among all consecutive slices within a volume in an unsupervised manner. This dynamic loss is based on curriculum learning to gradually learn information within a volume from local to global contexts. Additionally, the Inter-Slice Smoothness (ISS) loss is introduced as a regularization term to encourage changes between the slices occur consistently and continuously. During inference, Flow2Mask leverages these 3D flow fields for inter-slice mask propagation in a 3D image, spreading annotation from a single annotated slice to the entire volume. Moreover, we propose an automatic strategy to select the most representative slice as initial annotation in the mask propagation process. Experimental evaluations on different abdominal datasets demonstrate that our proposed SMP method outperforms previous approaches and improves the overall mean DSC of Vol2Flow by +2.1%, +8.2%, and +4.0% for the Sliver, CHAOS, and 3D-IRCAD datasets, respectively. Furthermore, Flow2Mask even exhibits substantial improvements in weakly-supervised and self-supervised few-shot segmentation methods when applied as a mask completion tool.

Normative representation learning focuses on understanding the typical anatomical distributions from large datasets of medical scans from healthy individuals. Generative Artificial Intelligence (AI) leverages this attribute to synthesize images that accurately reflect these normative patterns. This capability enables the AI allowing them to effectively detect and correct anomalies in new, unseen pathological data without the need for expert labeling. Traditional anomaly detection methods often evaluate the anomaly detection performance, overlooking the crucial role of normative learning. In our analysis, we introduce novel metrics, specifically designed to evaluate this facet in AI models. We apply these metrics across various generative AI frameworks, including advanced diffusion models, and rigorously test them against complex and diverse brain pathologies. In addition, we conduct a large multi-reader study to compare these metrics to experts’ evaluations. Our analysis demonstrates that models proficient in normative learning exhibit exceptional versatility, adeptly detecting a wide range of unseen medical conditions.

A major uncertainty in predicting the behaviour of marine-terminating glaciers is ice dynamics driven by non-linear calving front retreat, which is poorly understood and modelled. Using 124919 calving front positions for 149 marine-terminating glaciers in Svalbard from 1985 to 2023, generated with deep learning, we identify pervasive calving front retreats for non-surging glaciers over the past 38 years. We observe widespread seasonal cycles in calving front position for over half of the glaciers. At the seasonal timescale, peak retreat rates exhibit a several-month phase lag, with changes on the west coast occurring before those on the east coast, coincident with regional ocean warming. This spatial variability in seasonal patterns is linked to different timings of warm ocean water inflow from the West Spitsbergen Current, demonstrating the dominant role of ice-ocean interaction in seasonal front changes. The interannual variability of calving front retreat shows a strong sensitivity to both atmospheric and oceanic warming, with immediate responses to large air and ocean temperature anomalies in 2016 and 2019, likely driven by atmospheric blocking that can influence extreme temperature variability. With more frequent blocking occurring and continued regional warming, future calving front retreats will likely intensify, leading to more significant glacier mass loss.

Many scientific datasets are compositional in nature. Important biological examples include species abundances in ecology, cell-type compositions derived from single-cell sequencing data, and amplicon abundance data in microbiome research. Here, we provide a causal view on compositional data in an instrumental variable setting where the composition acts as the cause. First, we crisply articulate potential pitfalls for practitioners regarding the interpretation of compositional causes from the viewpoint of interventions and warn against attributing causal meaning to common summary statistics such as diversity indices in microbiome data analysis. We then advocate for and develop multivariate methods using statistical data transformations and regression techniques that take the special structure of the compositional sample space into account while still yielding scientifically interpretable results. In a comparative analysis on synthetic and real microbiome data we show the advantages and limitations of our proposal. We posit that our analysis provides a useful framework and guidance for valid and informative cause-effect estimation in the context of compositional data.

The ice thickness of the world’s glaciers is mostly unmeasured, and physics-based models to reconstruct ice thickness cannot always deliver accurate estimates. In this study, we use deep learning paired with physical knowledge to generate ice thickness estimates for all glaciers of Spitsbergen, Barentsøya, and Edgeøya in Svalbard. We incorporate mass conservation and other physically derived conditions into a neural network to predict plausible ice thicknesses even for glaciers without any in situ ice thickness measurements. With a glacier-wise cross-validation scheme, we evaluate the performance of the physics-informed neural network. The results of these proof-of-concept experiments let us identify several challenges and opportunities that affect the model’s performance in a real-world setting.

Active learning (AL) has shown promise to be a particularly data-efficient machine learning approach. Yet, its performance depends on the application, and it is not clear when AL practitioners can expect computational savings. Here, we carry out a systematic AL performance assessment for three diverse molecular datasets and two common scientific tasks: compiling compact, informative datasets and targeted molecular searches. We implemented AL with Gaussian processes (GP) and used the many-body tensor as molecular representation. For the first task, we tested different data acquisition strategies, batch sizes, and GP noise settings. AL was insensitive to the acquisition batch size, and we observed the best AL performance for the acquisition strategy that combines uncertainty reduction with clustering to promote diversity. However, for optimal GP noise settings, AL did not outperform the randomized selection of data points. Conversely, for targeted searches, AL outperformed random sampling and achieved data savings of up to 64%. Our analysis provides insight into this task-specific performance difference in terms of target distributions and data collection strategies. We established that the performance of AL depends on the relative distribution of the target molecules in comparison to the total dataset distribution, with the largest computational savings achieved when their overlap is minimal.