Active Learning Algorithm Research Articles

BSMArt: Simple and fast parameter space scans

We introduce BSMArt, a python program for the exploration of parameter spaces of theories Beyond the Standard Model. Especially designed for use with the SARAH family of tools, it is also sufficiently flexible to be used with a wide variety of external codes. BSMArt contains the first public release of the Active Learning scan by the same authors; but contains several additional scanning algorithms, ranging from the very simple to MultiNest and Diver. A BSMArt scan can be set up in a matter of minutes with only minimal editing of configuration files; installation scripts for all relevant tools and examples are provided. Program summaryProgram Title: BSMArtCPC Library link to program files:https://doi.org/10.17632/f6kwjvg8gw.1Developer's repository link:https://goodsell.pages.in2p3.fr/bsmart/Licensing provisions: GNU General Public License 3Programming language: pythonNature of problem: Exploring the parameter spaces of new physics models involves setting up, compiling and running multiple codes together. Then an algorithm must be chosen to select which parameter points to inspect in the simplest and quickest manner. Further, this code is required for the active machine learning algorithm by the same authors.Solution method: Automatic creation and compilation of the codes integrated with SARAH, and simple creation and management of input and output files allowing multithreaded running. Use of python facilitates machine learning applications. Multiple search algorithms are included including original ones.

Active Learning Guided Computational Discovery of Plant-Based Redoxmers for Organic Nonaqueous Redox Flow Batteries.

Organic nonaqueous redox flow batteries (O-NRFBs) are promising energy storage devices due to their scalability and reliance on sourceable materials. However, finding suitable redox-active organic molecules (redoxmers) for these batteries remains a challenge. Using plant-based compounds as precursors for these redoxmers can decrease their costs and environmental toxicity. In this computational study, flavonoid molecules have been examined as potential redoxmers for O-NRFBs. Flavone and isoflavone derivatives were selected as catholyte (positive charge carrier) and anolyte (negative charge carrier) molecules, respectively. To drive their redox potentials to the opposite extremes, in silico derivatization was performed using a novel algorithm to generate a library of > 40000 candidate molecules that penalizes overly complex structures. A multiobjective Bayesian optimization based active learning algorithm was then used to identify best redoxmer candidates in these search spaces. Our study provides methodologies for molecular design and optimization of natural scaffolds and highlights the need of incorporating expert chemistry awareness of the natural products and the basic rules of synthetic chemistry in machine learning.

SeeSaw: Interactive Ad-hoc Search Over Image Databases

As image datasets become ubiquitous, the problem of ad-hoc searches over image data is increasingly important. Many high-level data tasks in machine learning, such as constructing datasets for training and testing object detectors, imply finding ad-hoc objects or scenes within large image datasets as a key sub-problem. New foundational visual-semantic embeddings trained on massive web datasets such as Contrastive Language-Image Pre-Training (CLIP) can help users start searches on their own data, but we find there is a long tail of queries where these models fall short in practice. Seesaw is a system for interactive ad-hoc searches on image datasets that integrates state-of-the-art embeddings like CLIP with user feedback in the form of box annotations to help users quickly locate images of interest in their data even in the long tail of harder queries. One key challenge for Seesaw is that, in practice, many sensible approaches to incorporating feedback into future results, including state-of-the-art active-learning algorithms, can worsen results compared to introducing no feedback, partly due to CLIP's high-average performance. Therefore, Seesaw includes several algorithms that empirically result in larger and also more consistent improvements. We compare Seesaw's accuracy to both using CLIP alone and to a state-of-the-art active-learning baseline and find Seesaw consistently helps improve results for users across four datasets and more than a thousand queries. Seesaw increases Average Precision (AP) on search tasks by an average of .08 on a wide benchmark (from a base of .72), and by a .27 on a subset of more difficult queries where CLIP alone performs poorly.

Atomistic fracture in bcc iron revealed by active learning of Gaussian approximation potential

The prediction of atomistic fracture mechanisms in body-centred cubic (bcc) iron is essential for understanding its semi-brittle nature. Existing atomistic simulations of the crack-tip under mode-I loading based on empirical interatomic potentials yield contradicting predictions and artificial mechanisms. To enable fracture prediction with quantum accuracy, we develop a Gaussian approximation potential (GAP) using an active learning strategy by extending a density functional theory (DFT) database of ferromagnetic bcc iron. We apply the active learning algorithm and obtain a Fe GAP model with a converged model uncertainty over a broad range of stress intensity factors (SIFs) and for four crack systems. The learning efficiency of the approach is analysed, and the predicted critical SIFs are compared with Griffith and Rice theories. The simulations reveal that cleavage along the original crack plane is the atomistic fracture mechanism for {100} and {110} crack planes at T = 0 K, thus settling a long-standing issue. Our work also highlights the need for a multiscale approach to predicting fracture and intrinsic ductility, whereby finite temperature, finite loading rate effects and pre-existing defects (e.g., nanovoids, dislocations) should be taken explicitly into account.

Multi-principal element alloy discovery using directed energy deposition and machine learning

Multi-principal element alloys open large composition spaces for alloy development. The large compositional space necessitates rapid synthesis and characterization to identify promising materials, as well as predictive strategies for alloy design. Additive manufacturing via directed energy deposition is demonstrated as a high-throughput technique for synthesizing alloys in the Cr-Fe-Mn-Ni quaternary system. More than 100 compositions are synthesized in a week, exploring a broad range of compositional space. Uniform compositional control to within ±5 at% is achievable. The rapid synthesis is combined with conjoint sample heat treatment (25 samples vs 1 sample), and automated characterization including X-ray diffraction, energy-dispersive X-ray spectroscopy, and nano-hardness measurements. The datasets of measured properties are then used for a predictive strengthening model using an active machine learning algorithm that balances exploitation and exploration. A learned parameter that represents lattice distortion is trained using the alloy compositions. This combination of rapid synthesis, characterization, and active learning model results in new alloys that are significantly stronger than previous investigated alloys.

Calibration of uncertainty in the active learning of machine learning force fields

FFLUX is a machine learning force field that uses the maximum expected prediction error (MEPE) active learning algorithm to improve the efficiency of model training. MEPE uses the predictive uncertainty of a Gaussian process (GP) to balance exploration and exploitation when selecting the next training sample. However, the predictive uncertainty of a GP is unlikely to be accurate or precise immediately after training. We hypothesize that calibrating the uncertainty quantification within MEPE will improve active learning performance. We develop and test two methods to improve uncertainty estimates: post-hoc calibration of predictive uncertainty using the CRUDE algorithm, and replacing the GP with a student-t process. We investigate the impact of these methods on MEPE for single sample and batch sample active learning. Our findings suggest that post-hoc calibration does not improve the performance of active learning using the MEPE method. However, we do find that the student-t process can outperform active learning strategies and random sampling using a GP if the training set is sufficiently large.

Exploiting redundancy in large materials datasets for efficient machine learning with less data

Extensive efforts to gather materials data have largely overlooked potential data redundancy. In this study, we present evidence of a significant degree of redundancy across multiple large datasets for various material properties, by revealing that up to 95% of data can be safely removed from machine learning training with little impact on in-distribution prediction performance. The redundant data is related to over-represented material types and does not mitigate the severe performance degradation on out-of-distribution samples. In addition, we show that uncertainty-based active learning algorithms can construct much smaller but equally informative datasets. We discuss the effectiveness of informative data in improving prediction performance and robustness and provide insights into efficient data acquisition and machine learning training. This work challenges the “bigger is better” mentality and calls for attention to the information richness of materials data rather than a narrow emphasis on data volume.

DivTheft: An Ensemble Model Stealing Attack by Divide-and-Conquer

Recently, model stealing attacks are widely studied but most of them are focused on stealing a single non-discrete model, e.g., neural networks. For ensemble models, these attacks are either non-executable or suffer from intolerant performance degradation due to the complex model structure (multiple sub-models) and the discreteness possessed by the sub-model (e.g., decision trees). To overcome the bottleneck, this paper proposes a divide-and-conquer strategy called DivTheft to formulate the model stealing attack to common ensemble models by combining active learning (AL). Specifically, based on the boosting learning concept, we divide a hard ensemble model stealing task into multiple simpler ones about single sub-model stealing. Then, we adopt AL to conquer the data-free sub-model stealing task. During the process, the current AL algorithm easily causes the stolen model to be biased because of ignoring the past useful memories. Thus, DivTheft involves a newly designed uncertainty sampling scheme to filter reusable samples from the previously used ones. Experiments show that compared with the prior work, DivTheft can save almost 50% queries while ensuring a competitive agreement rate to the victim model.

Efficient Exploration of Chemical Compound Space Using Active Learning for Prediction of Thermodynamic Properties of Alkane Molecules.

We introduce an exploratory active learning (AL) algorithm using Gaussian process regression and marginalized graph kernel (GPR-MGK) to sample chemical compound space (CCS) at minimal cost. Targeting 251,728 enumerated alkane molecules with 4-19 carbon atoms, we applied the AL algorithm to select a diverse and representative set of molecules and then conducted high-throughput molecular simulations on these selected molecules. To demonstrate the power of the AL algorithm, we built directed message-passing neural networks (D-MPNN) using simulation data as the training set to predict liquid densities, heat capacities, and vaporization enthalpies of the CCS. Validations show that D-MPNN models built on the smallest training set considered in this work, which consists of 313 molecules or 0.124% of the original CCS, predict the properties with R2 > 0.99 against the computational data and R2 > 0.94 against the experimental data. The advantage of the presented AL algorithm is that the predicted uncertainty of GPR depends on only the molecular structures, which renders it compatible with high-throughput data generation.

Open set transfer learning through distribution driven active learning

Domain adaptation enables effective transfer between source and target domains with different distributions. The latest research focuses on open set domain adaptation; that is, the target domain contains unknown categories that do not exist in the source domain. The existing open set domain adaptation cannot realize the fine-grained recognition of unknown categories. In this paper, we propose an uncertainty analysis evidence model and design a distribution driven active transfer learning (DATL) algorithm. DATL realizes fine-grained recognition of unknown categories with no requirements on the source domain to contain the unknown categories. To explore unknown distributions, the uncertainty analysis evidence model was adopted to divide the high uncertainty space. To select critical instances, a cluster-diversity query strategy was proposed to identify new categories. To enrich the label categories of the source domain, a global dynamic alignment strategy was designed to avoid negative transfers. Comparative experiments with state-of-the-art methods on the standard Office-31/Office-Home/Office-Caltech10 benchmarks showed that the DATL algorithm: (1) outperformed its competitors; (2) realized accurate identification of unknown subcategories from a fine-grained perspective; and (3) achieved outstanding performance even with a very high degree of openness.

Online Passive-Aggressive Active Learning for Trapezoidal Data Streams.

The idea of combining the active query strategy and the passive-aggressive (PA) update strategy in online learning can be credited to the PA active (PAA) algorithm, which has proven to be effective in learning linear classifiers from datasets with a fixed feature space. We propose a novel family of online active learning algorithms, named PAA learning for trapezoidal data streams (PAA [Formula: see text] ) and multiclass PAA [Formula: see text] (MPAA [Formula: see text] ) (and their variants), for binary and multiclass online classification tasks on trapezoidal data streams where the feature space may expand over time. Under the context of an ever-changing feature space, we provide the theoretical analysis of the mistake bounds for both PAA [Formula: see text] and MPAA [Formula: see text] . Our experiments on a wide variety of benchmark datasets have confirm that the combination of the instance-regulated active query strategy and the PA update strategy is much more effective in learning from trapezoidal data streams. We have also compared PAA [Formula: see text] with online learning with streaming features (OL [Formula: see text] )-the state-of-the-art approach in learning linear classifiers from trapezoidal data streams. PAA [Formula: see text] could achieve much better classification accuracy, especially for large-scale real-world data streams.

Multiclass autoencoder-based active learning for sensor-based human activity recognition

Leveraging the enormous amounts of real-world data collected through Internet of Things (IoT) technologies, human activity recognition (HAR) has become a crucial component of numerous human-centric applications, with the aim of enhancing the quality of human life. While the recent advancements in deep learning have significantly improved HAR, the process of labeling data continues to remain a significant challenge due to the substantial costs associated with human annotation for supervised model training. Active learning (AL) addresses this issue by strategically selecting informative samples for labeling during model training, thereby enhancing model performance. Although numerous approaches have been proposed for sample selection, which consider aspects of uncertainty and representation, the difficulties in estimating uncertainty and exploiting distribution of high-dimensional data still pose a major issue. Our proposed deep learning-based active learning algorithm, called Multiclass Autoencoder-based Active Learning (MAAL), learns latent representation leveraging the capacity of Deep Support Vector Data Description (Deep SVDD). With the multiclass autoencoder which learns the normal characteristics of each activity class in the latent space, MAAL provides an informative sample selection for model training by establishing a link between the HAR model and the selection model. We evaluate our proposed MAAL using two publicly available datasets. The performance results demonstrate the improvements across the overall active learning rounds, achieving enhancements up to 3.23% accuracy and 3.67% in the F1 score. Furthermore, numerical results and analysis of sample selection are presented to validate the effectiveness of the proposed MAAL compared to the alternative comparison methods.

PLVI-CE: a multi-label active learning algorithm with simultaneously considering uncertainty and diversity

PLVI-CE: a multi-label active learning algorithm with simultaneously considering uncertainty and diversity

Synthesis-Driven Computational Discovery of Organic Redoxmers for Non-Aqueous Redox Flow Batteries

Organic non-aqueous redox flow batteries (RFBs) are promising grid-scale energy storage systems for storing intermittent renewable energy in molecules. For practical applications, low-cost and stable redox active molecules (redoxmers) that display a large redox potential window and long electrochemical cycling stability are required to deliver a high-energy-density RFB with a long battery cycle life. To accelerate the discovery of redoxmer molecules, previous studies have utilized machine learning (ML) methods along with first-principles simulations. Although many ML-suggested molecules show promising electrochemical properties, they are generally difficult to synthesize because of their highly complex structure. To address this challenge, in this work, we propose a new synthesis-driven computational approach to discovering less-complex molecules with a large redox potential window that can deliver high-energy-density RFBs. Specifically, we devised an algorithm to generate an engineered molecular search space and applied a Bayesian optimization-based active learning algorithm to discover promising molecules from a large molecular search space while using a limited number of expensive density functional theory (DFT) calculations. To increase the energy density of RFBs, we focus on maximizing the redox potential window of redoxmer molecules (or cell voltage) by minimizing the reduction potential of anolyte molecules. Precisely, we studied the 2,1,3 Benzothiadiazole (BTZ) molecule, a promising anolyte (redoxmer) candidate for RFBs, and with our synthesis-driven approach discovered new BTZ-based molecules that show 15-43% smaller reduction potential than the BTZ and in principle can deliver organic RFBs with a cell voltage ≈ 3 V.

Atom-centered machine-learning force field package

In recent years, machine learning algorithms have been widely used for constructing force fields with an accuracy of ab initio methods and the efficiency of classical force fields. Here, we developed a python-based atom-centered machine-learning force field (PyAMFF) package to provide a simple and efficient platform for fitting and using machine learning force fields by implementing an atom-centered neural-network algorithm with Behler-Parrinello symmetry functions as structural fingerprints. The following three features are included in PyAMFF: (1) integrated Fortran modules for fast fingerprint calculations and Python modules for user-friendly integration through scripts and facile extension of future algorithms; (2) a pure Fortran backend to interface with the software, including the long-timescale dynamic simulation package EON, enabling both molecular dynamic simulations and adaptive kinetic Monte Carlo simulations with machine-learning force fields; and (3) integration with the Atomic Simulation Environment package for active learning and ML-based algorithm development. Here, we demonstrate an efficient parallelization of PyAMFF in terms of CPU and memory usage and show that the Fortran-based PyAMFF calculator exhibits a linear scaling relationship with the number of symmetry functions and the system size. Program summaryProgram title: python-based atom-centered machine-learning force field (PyAMFF)CPC Library link to program files:https://doi.org/10.17632/fsn6dkcvrv.1Developer's repository link:https://gitlab.com/pyamff/pyamffLicensing provisions: Apache License, 2.0Nature of problem: Determine an approximate (surrogate) model based upon atomic forces and energies from density functional theory (DFT). With a surrogate model that is less computationally expensive to evaluate than DFT, there can be a rapid exploration of the potential energy surface, accelerated optimization to minima and saddle points, and ultimately, accelerated design of active materials where the kinetics are key to the material function.Solution method: The atomic environments of training data are calculated in terms of Behler-Parrinello fingerprints. These fingerprints are passed to a neural network which is trained to reproduce the energy and force of the training data. A parallel implementation and Fortran backend allow for efficient training and calculation of the resulting surrogate model. Examples of long-time simulations of materials on the surrogate model surfaces are provided.

Active Machine Learning for Chemical Engineers: A Bright Future Lies Ahead!

By combining machine learning with the design of experiments, thereby achieving so-called active machine learning, more efficient and cheaper research can be conducted. Machine learning algorithms are more flexible and are better than traditional design of experiment algorithms at investigating processes spanning all length scales of chemical engineering. While active machine learning algorithms are maturing, their applications are falling behind. In this article, three types of challenges presented by active machine learning—namely, convincing the experimental researcher, the flexibility of data creation, and the robustness of active machine learning algorithms—are identified, and ways to overcome them are discussed. A bright future lies ahead for active machine learning in chemical engineering, thanks to increasing automation and more efficient algorithms that can drive novel discoveries.

Towards active learning: A stopping criterion for the sequential sampling of grain boundary degrees of freedom

Many materials processes and properties depend on the anisotropy of the energy of grain boundaries, i.e. on the fact that this energy is a function of the five geometric degrees of freedom (DOF) of the interface. To access this parameter space in an efficient way and to discover energy cusps in unexplored regions, a method was recently established, which combines atomistic simulations with statistical methods (Kroll et al., 2022). This sequential sampling technique is now extended in the spirit of an active learning algorithm by adding a criterion to decide when the sampling has advanced enough to stop. In this instance, two parameters to analyse the sampling results on the fly are introduced: the number of cusps, which correspond to the most interesting and important regions of the energy landscape, and the maximum change of energy between two sequential iterations. Monitoring these two quantities provides valuable insight into how the subspaces are energetically structured. The combination of both parameters provides the necessary information to evaluate the sampling of the 2D subspaces of grain boundary plane inclinations of even non-periodic, low angle grain boundaries. With a reasonable number of data points in the initial design, only a few appropriately chosen sequential iterations already improve the accuracy of the sampling substantially and unknown cusps can be found within a few additional sequential steps.

Inversion of maize leaf area index from UAV hyperspectral and multispectral imagery

The accurate estimationof Leaf area index (LAI)is of great importance for evaluating crop growth in precision agriculture. Although previous studies have confirmed great advantages ofunmanned aerial vehicle(UAV) remote sensing for LAI estimation in the field, accurate, reliable and efficient LAI estimation with practical applications still faces challenges due to model limitations and variations in the spectral and spatial scales of UAV remote sensing. In this study, we constructed the hybrid inversion models (HIMs) for estimating maize LAI using UAV hyperspectral and multispectral data, respectively. The HIMs combines the advantages of radiative transfer models and machine learning regression algorithms, and are optimized by active learning (AL) algorithm. The results reveal that the inclusion of AL in the HIMs an effectively improve the accuracy of the model. The Gaussian process regression-based HIM with AL (GPR-AL-HIM) obtained the best performance in LAI estimation (R2 = 0.86, RMSE = 0.30 and NRMSE = 10.16 %). Meanwhile, GPR-AL-HIM was also determined to outperform the physical model based on the look-up-table (LUT) and the empirical statistical model based on vegetation indices. The model was validated with another independent dataset and also obtained a high accuracy (R2 = 0.84, RMSE = 0.23 and NRMSE = 11.78 %). In addition, we also explore the effects of the UAV spectral (multispectral and hyperspectral) and image spatial resolution on LAI inversion. The results reveal that the hyperspectral data exhibit an advantage over the multispectral data for LAI inversion using the GPR-AL-HIM. The accuracy of the GPR-AL-HIM decreased with increasing spatial resolution, but the accuracy varied less within a certain spatial resolution range (e.g., R2of 0.86–0.84 and RMSE of 0.30–0.33 for hyperspectral images at 1.5–15 cm spatial resolution). Furthermore, the LAI distribution in the study area was accurately mapped using the GPR-AL-HIM with the hyperspectral and multispectral images, with the latter exhibiting lower uncertainties. The GPR-AL-HIM is mainly aimed at maize, and in the future, we will explore the applicability of this model in other crops. This work provides a reference for the design of a monitoring scheme with crop parameters based on UAV remote sensing in precision agriculture.

Research on classification method of high myopic maculopathy based on retinal fundus images and optimized ALFA-Mix active learning algorithm.

To conduct a classification study of high myopic maculopathy (HMM) using limited datasets, including tessellated fundus, diffuse chorioretinal atrophy, patchy chorioretinal atrophy, and macular atrophy, and minimize annotation costs, and to optimize the ALFA-Mix active learning algorithm and apply it to HMM classification. The optimized ALFA-Mix algorithm (ALFA-Mix+) was compared with five algorithms, including ALFA-Mix. Four models, including ResNet18, were established. Each algorithm was combined with four models for experiments on the HMM dataset. Each experiment consisted of 20 active learning rounds, with 100 images selected per round. The algorithm was evaluated by comparing the number of rounds in which ALFA-Mix+ outperformed other algorithms. Finally, this study employed six models, including EfficientFormer, to classify HMM. The best-performing model among these models was selected as the baseline model and combined with the ALFA-Mix+ algorithm to achieve satisfactory classification results with a small dataset. ALFA-Mix+ outperforms other algorithms with an average superiority of 16.6, 14.75, 16.8, and 16.7 rounds in terms of accuracy, sensitivity, specificity, and Kappa value, respectively. This study conducted experiments on classifying HMM using several advanced deep learning models with a complete training set of 4252 images. The EfficientFormer achieved the best results with an accuracy, sensitivity, specificity, and Kappa value of 0.8821, 0.8334, 0.9693, and 0.8339, respectively. Therefore, by combining ALFA-Mix+ with EfficientFormer, this study achieved results with an accuracy, sensitivity, specificity, and Kappa value of 0.8964, 0.8643, 0.9721, and 0.8537, respectively. The ALFA-Mix+ algorithm reduces the required samples without compromising accuracy. Compared to other algorithms, ALFA-Mix+ outperforms in more rounds of experiments. It effectively selects valuable samples compared to other algorithms. In HMM classification, combining ALFA-Mix+ with EfficientFormer enhances model performance, further demonstrating the effectiveness of ALFA-Mix+.

Association between contrast sensitivity function and structural damage in primary open-angle glaucoma

AimsTo evaluate the association between contrast sensitivity function (CSF) and glaucomatous structural damage in primary open-angle glaucoma (POAG).MethodsA cross-sectional study was performed with 103 patients (103 eyes) aged 25–50 years...