Black-box Algorithms Research Articles

Zero-Shot Discovery of High-Performance, Low-Cost Organic Battery Materials Using Machine Learning.

Organic electrode materials (OEMs), composed of abundant elements such as carbon, nitrogen, and oxygen, offer sustainable alternatives to conventional electrode materials that depend on finite metal resources. The vast structural diversity of organic compounds provides a virtually unlimited design space; however, exploring this space through Edisonian trial-and-error approaches is costly and time-consuming. In this work, we develop a new framework, SPARKLE, that combines computational chemistry, molecular generation, and machine learning to achieve zero-shot predictions of OEMs that simultaneously balance reward (specific energy), risk (solubility), and cost (synthesizability). We demonstrate that SPARKLE significantly outperforms alternative black-box machine learning algorithms on interpolation and extrapolation tasks. By deploying SPARKLE over a design space of more than 670,000 organic compounds, we identified ≈5000 novel OEM candidates. Twenty-seven of them were synthesized and fabricated into coin-cell batteries for experimental testing. Among SPARKLE-discovered OEMs, 62.9% exceeded benchmark performance metrics, representing a 3-fold improvement over OEMs selected by human intuition alone (20.8% based on six years of prior lab experience). The top-performing OEMs among the 27 candidates exhibit specific energy and cycling stability that surpass the state-of-the-art while being synthesizable at a fraction of the cost.

Discretization-independent node-based shape optimization with the Vertex Morphing method using design variable scaling

AbstractThe Vertex Morphing method is a node-based shape parameterization that uses an explicit filtering approach to regularize the optimization problem and generate smooth shapes. It has been successfully applied to shape optimization problems of industrial size in recent years. This work investigates in detail how irregular discretizations, design surface boundaries, and complex geometries can influence the progress of a gradient-based optimization using the standard Vertex Morphing formulation. A sensitivity weighting approach based on the available shape morphing functions is presented, which eliminates all of the aforementioned influences. Subsequently, a design variable scaling strategy is developed that transforms the optimization problem into an alternative design space and allows the use of arbitrary, even highly irregular surface discretizations in combination with black-box optimization algorithms for shape optimization with the Vertex Morphing method. Illustrative academic examples and an application case of an additively manufactured part are presented to support the work.

Combining categorical boosting and Shapley additive explanations for building an interpretable ensemble classifier for identifying mineralization-related geochemical anomalies

The vast majority of shallow and deep learning techniques used to identify mineralization-related geochemical anomalies are black-box algorithms that lack the ability to elucidate the individual contributions of each element towards the model predictions. In addition, most of the anomaly identification models established by both shallow and deep learning algorithms lack robustness. Establishing interpretable and robust machine learning models is a challenge in applying machine learning techniques to geochemical anomaly identification. To this end, the categorical boosting (CatBoost) algorithm was employed to build a robust ensemble classifier to identify mineralization-related anomalies from the 1:50,000 geochemical reconnaissance data (stream sediment survey) in the Yeniugou area of Xinjiang (China). The receiver operating characteristic curve (ROC) and precision-recall (P-R) curve of the ensemble model were plotted, and the area under the ROC curve (AUC) as well as the area under the P-R curve (AUPRC) of the ensemble model were calculated to measure the performance of the ensemble model. The ROC curve of the ensemble model approximates that of the perfect classification model. The P-R curve of the ensemble model is close to the upper right corner of the P-R space. The AUC and AUPRC values of the ensemble model reaches 0.9981 and 0.7816, respectively. The identified polymetallic mineralization-related geochemical anomalies account for 3% of the whole exploration area, correctly identifying all known polymetallic deposits. To enhance the interpretability of the CatBoost model, the Shapley additive explanations (SHAP) tool was adopted to graphically interpret the predictions of the ensemble model. The graphic interpretation shows that the importance order of the 14 elements is Ni-Au-Ag-Sn-As-Cr-Zn-Cu-Pb-Sb-W-Bi-Mo-Co. Cu and Ni are most likely metallogenic elements of the study area. Cu interacts with Ni, Ag, As, Sn, Cr, Zn, Pb, Sb, W, Bi, and Co; and Ni interacts with Au, Sn, As, Zn, Cu, W, Bi, and Co. Two polymetallic prospective areas were delineated in the study area. One is Cu-Ni-polymetallic mineralization prospective area, and the other is Ni-polymetallic mineralization prospective area. It can be concluded that the combination of CatBoost and SHAP is an effective way to construct an interpretable ensemble model with high-performance and robustness in identifying mineralization-related geochemical anomalies.

Attention Heat Map-Based Black-Box Local Adversarial Attack for Synthetic Aperture Radar Target Recognition

Synthetic aperture radar (SAR) automatic target recognition (ATR) models based on deep neural networks (DNNs) are susceptible to adversarial attacks. In this study, we proposed an SAR black-box local adversarial attack algorithm named attention heat map- based black-box local adversarial attack (AH-BLAA). First, we designed an attention heat map extraction module combined with the layer-wise relevance propagation (LRP) algorithm to obtain the high concerning areas of the SAR-ATR models. Then, to gener- ate SAR adversarial attack examples, we designed a perturbation generator module, introducing the structural dissimilarity (DSSIM) metric in the loss function to limit image distortion and the dif- ferential evolution (DE) algorithm to search for optimal perturba- tions. Experimental results on the MSTAR and FUSAR-Ship datasets showed that compared with existing adversarial attack algorithms, the attack success rate of the AH-BLAA algorithm increased by 0.63% to 33.59% and 1.05% to 17.65%, respectively. Moreover, the low- est perturbation ratios reached 0.23% and 0.13%, respectively.

Evaluation of stress distributions in trimaterial bonded joints with nano-resin adhesive using machine learning models

Adhesive bonded joints hold significant importance across various industrial sectors in modern engineering, owing to their lightweight nature and myriad advantages. The rising demand for trimaterial joints underscores their utility and versatility. In these joints, the choice of materials for both adherends greatly influences their strength, structural reliability, and overall characteristics. While numerous researches have extensively analyzed stress distributions, their effects, and behaviors, many have relied on a one-factor-at-a-time approach, focusing solely on individual design variables' effects. However, recognizing the intricate interplay among various material combinations and their collective impact on overall performance, this study employs various types of White-box, Black-box, and Grey-box machine learning algorithms to identify an optimized ML model as well as predict stress distributions for any random combinations of upper and lower adherend materials. Dataset of total 178 random material combinations were utilized for the training phases with 5-fold cross validation and model tuning. However, the decision tree regressor emerged as the optimized model by comparing the quantitative metrics of accuracy benchmark as well as the prediction outcomes obtained through all the machine learning models. The maximum prediction accuracy attained was an impressive 99.97 %, while the minimum recorded was 89.74 %. This research aims to identify tailored machine learning model specifically for trimaterial bonded joints where nano layer of resin is utilized as the adhesive.

Using Machine Learning Methods to Assess Module Performance Contribution in Modular Optimization Frameworks.

Modular algorithm frameworks not only allow for combinations never tested in manually selected algorithm portfolios, but they also provide a structured approach to assess which algorithmic ideas are crucial for the observed performance of algorithms. In this study, we propose a methodology for analyzing the impact of the different modules on the overall performance. We consider modular frameworks for two widely used families of derivative-free black-box optimization algorithms, the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) and differential evolution (DE). More specifically, we use performance data of 324 modCMA-ES and 576 modDE algorithm variants (with each variant corresponding to a specific configuration of modules) obtained on the 24 BBOB problems for 6 different runtime budgets in 2 dimensions. Our analysis of these data reveals that the impact of individual modules on overall algorithm performance varies significantly. Notably, among the examined modules, the elitism module in CMA-ES and the linear population size reduction module in DE exhibit the most significant impact on performance. Furthermore, our exploratory data analysis of problem landscape data suggests that the most relevant landscape features remain consistent regardless of the configuration of individual modules, but the influence that these features have on regression accuracy varies. In addition, we apply classifiers that exploit feature importance with respect to the trained models for performance prediction and performance data, to predict the modular configurations of CMA-ES and DE algorithm variants. The results show that the predicted configurations do not exhibit a statistically significant difference in performance compared to the true configurations, with the percentage varying depending on the setup (from 49.1% to 95.5% for mod-CMA and 21.7% to 77.1% for DE).

Intrinsically interpretable machine learning-based building energy load prediction method with high accuracy and strong interpretability

Black-box models have demonstrated remarkable accuracy in forecasting building energy loads. However, they usually lack interpretability and do not incorporate domain knowledge, making it difficult for users to trust their predictions in practical applications. One important and interesting question remains unanswered: is it possible to use intrinsically interpretable models to achieve accuracy comparable to that of black-box models? With an aim of answering this question, this study proposes an intrinsically interpretable machine learning-based method to forecast building energy loads. It creatively combines two intrinsically interpretable machine learning algorithms: clustering decision trees and adaptive multiple linear regression. Clustering decision trees aim to automatically identify various building operation conditions, allowing for the training of multiple models tailored to each condition. It can reduce the complexity of model training data, leading to higher accuracy. Adaptive multiple linear regression is an improved regression algorithm tailored to building energy load prediction. It can adaptively modify regression coefficients according to building operations, enhancing the non-linear fitting capability of multiple linear regression. The proposed method is evaluated utilizing the operational data from an office building. The results indicate that the proposed method exhibits comparable accuracy to both random forests and extreme gradient boosting. Furthermore, it shows significantly superior accuracy, with an average improvement of 10.2 %, compared with some popular black-box algorithms such as artificial neural networks, support vector regression, and classification and regression trees. As for model interpretability, the proposed method reveals that historical cooling loads are the most crucial for predicting building cooling loads under most conditions. Additionally, outdoor air temperature has a significant contribution to building cooling load prediction during the daytime on weekdays in summer and transition seasons. In the future, it will be valuable to explore integrating the laws of physics into the proposed method to further enhance its interpretability.

Parameter estimation for Gibbs distributions

A central problem in computational statistics is to convert a procedure for sampling combinatorial objects into a procedure for counting those objects, and vice versa. We consider sampling problems coming from Gibbs distributions , which are families of probability distributions over a discrete space \(\Omega\) with probability mass function of the form \(\mu^{\Omega}_{\beta}(\omega)\propto e^{\beta H(\omega)}\) for \(\beta\) in an interval \([\beta_{\min},\beta_{\max}]\) and \(H(\omega)\in\{0\}\cup[1,n]\) . Two important parameters are the partition function , which is the normalization factor \(Z(\beta)=\sum_{\omega\in\Omega}e^{\beta H(\omega)}\) , and the vector of preimage counts \(c_{x}=|H^{-1}(x)|\) . We develop black-box sampling algorithms to estimate the counts using roughly \(\tilde{O}(\frac{n^{2}}{\varepsilon^{2}})\) samples for integer-valued distributions and \(\tilde{O}(\frac{q}{\varepsilon^{2}})\) samples for general distributions, where \(q=\log\frac{Z(\beta_{\max})}{Z(\beta_{\min})}\) (ignoring some second-order terms and parameters). We show this is optimal up to logarithmic factors. We illustrate with improved algorithms for counting connected subgraphs, independent sets, and perfect matchings. As a key subroutine, we estimate all values of the partition function using \(\tilde{O}(\frac{n^{2}}{\varepsilon^{2}})\) samples for integer-valued distributions and \(\tilde{O}(\frac{q}{\varepsilon^{2}})\) samples for general distributions. This improves over a prior algorithm of Huber (2015) which computes a single point estimate \(Z(\beta_{\max})\) and which uses a slightly larger amount of samples. We show matching lower bounds, demonstrating this complexity is optimal as a function of \(n\) and \(q\) up to logarithmic terms.

A hybrid machine learning strategy for pitting probability prediction of stainless steels

Pitting corrosion represents a substantial threat to passive metals, potentially leading to stress-corrosion cracking and the abrupt failure of materials. The complex relationship between the internal alloy compositions and external environmental factors complicates the precise prediction of pitting probability. In this study, a hybrid machine learning strategy is developed to predict the probability and boundary of pitting corrosion in stainless steel, considering the collaborative effects of different factor pairs rather than a single factor. The strategy integrates both black-box (adaptive boosting) and white-box (symbolic classification) algorithms, where the former is constructed to determine the pitting probability with an F1 score of 0.85, and the latter is designed to achieve recognition of critical factor pairs. This work aims to offer a systematic diagram and new insights in understanding and optimizing the pitting resistance of stainless steel.

Variety Evasive Subspace Families

We introduce the problem of constructing explicit variety evasive subspace families. Given a family F\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal{F}$$\\end{document} of subvarieties of a projective or affine space, a collection H\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal{H}$$\\end{document} of projective or affine k\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$k$$\\end{document}-subspaces is (F,ϵ)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$(\\mathcal{F},\\epsilon)$$\\end{document}-evasive iffor every V∈F\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal{V}\\in\\mathcal{F}$$\\end{document}, all but at most ϵ\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\epsilon$$\\end{document}-fraction of W∈H\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$W\\in\\mathcal{H}$$\\end{document} intersect every irreducible component of V\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mathcal{V}$$\\end{document} with (at most) the expected dimension.The problem of constructing such an explicit subspace family generalizes both deterministic black-box polynomial identity testing (PIT) and the problem of constructing explicit (weak) lossless rank condensers.Using Chow forms, we construct explicit k\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$k$$\\end{document}-subspace families of polynomial size that are evasive for all varieties of bounded degree in a projective or affine n\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$n$$\\end{document}-space. As one application, we obtain a complete derandomization of Noether’s normalization lemma for varieties of low degree in a projective or affine n\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$n$$\\end{document}-space. In another application, we obtain a simple polynomial-time black-box PIT algorithm for depth-4 arithmetic circuits with bounded top fan-in and bottom fan-in that are not in the Sylvester–Gallai configuration, improving and simplifying a result of Gupta (ECCC TR 14-130).As a complement of our explicit construction, we prove a tight lower bound for the size of k\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$k$$\\end{document}-subspace families that are evasive for degree-d\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$d$$\\end{document} varieties in a projective n\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$n$$\\end{document}-space. When n-k=nΩ(1)\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$n-k=n^{\\Omega(1)}$$\\end{document}, the lower bound is superpolynomial unless d\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$d$$\\end{document} is bounded. The proof uses a dimension counting argument on Chow varieties that parametrize projective subvarieties.

Exploring the decision-making process of ensemble learning algorithms in landslide susceptibility mapping: Insights from local and global explainable AI analyses

Artificial intelligence and machine learning have attracted significant attention in the preparation of landslide susceptibility maps (LSMs) over the years. Achieving considerable success, they frequently face criticism for their opaque nature and the limited capacity to explain and interpret the resulting LSMs. This study uncovers the inherent characteristics of conditioning factors by investigating both local and global driving forces influencing landslide events through the lens of explainable artificial intelligence. To accomplish this, black-box algorithms, including random forest, gradient boosting machines, and extreme gradient boosting, as well as white-box algorithms including logistic regression and decision trees, were employed to generate LSMs. Their internal structures were later illuminated with three global and one local explainable artificial intelligence (XAI) techniques. The results unveiled a significant superiority of black-box algorithms over white-box algorithms, demonstrating an improvement of up to 17 % in overall accuracy and 19 % in the area under the curve (AUC) score. Among them, the gradient boosting machines exhibited the highest performance, achieving an overall accuracy of 87.88 % and an AUC of 0.9382. Global explanation analyses revealed that landslide susceptibility was predominantly influenced by slope, elevation, distance to roads, and lithological units. On the other hand, local interpretations, conducted for three specific landslide cases, disclosed relative variations in the importance of causative factors such as slope and distance to rivers on landslide occurrences. Overall, this study illustrates the potential utility of XAI tools in enhancing the transparency of generated maps and elucidating the underlying causes of specific landslide occurrences.

Transferring Black-Box Decision Making to a White-Box Model

In the rapidly evolving realm of artificial intelligence (AI), black-box algorithms have exhibited outstanding performance. However, their opaque nature poses challenges in fields like medicine, where the clarity of the decision-making processes is crucial for ensuring trust. Addressing this need, the study aimed to augment these algorithms with explainable AI (XAI) features to enhance transparency. A novel approach was employed, contrasting the decision-making patterns of black-box and white-box models. Where discrepancies were noted, training data were refined to align a white-box model’s decisions closer to its black-box counterpart. Testing this methodology on three distinct medical datasets revealed consistent correlations between the adapted white-box models and their black-box analogs. Notably, integrating this strategy with established methods like local interpretable model-agnostic explanations (LIMEs) and SHapley Additive exPlanations (SHAPs) further enhanced transparency, underscoring the potential value of decision trees as a favored white-box algorithm in medicine due to its inherent explanatory capabilities. The findings highlight a promising path for the integration of the performance of black-box algorithms with the necessity for transparency in critical decision-making domains.

A quantitative investigation on pyrolysis behaviors of metal ion-exchanged coal macerals by interpretable machine learning algorithms

Generalizing the rules from complex processes such as catalytic pyrolysis to guide their process control is always a difficult but attractive task. The influences of ion-exchange of metal ions (Na+, K+, Ca2+, Mg2+, Co2+ and Ni2+) on the pyrolysis behavior of vitrinite and inertinite from Shendong coal were investigated by thermogravimetric analyzer-Fourier transform infrared spectrometer (TG-FTIR), fixed-bed reactor (FBR), gas chromatograph-mass spectrometer (GC-MS) and X-ray diffractometer (XRD). A set of machine learning models was successfully constructed based on random forest, support vector machine, and Gaussian process regression, to quantify the relationships between pyrolysis behaviors and the properties of metals and macerals. The leave-one-out-cross validation showed that there are considerable determination coefficients (R2 > 0.9) between predicted and experimental values for most responses (17 out of 29). By combining genetic programming-based symbolic regression with the black-box algorithm, 23 symbolic regression expressions with high confidence were successfully constructed. This work is a pioneering attempt of optimization in small-scale experiments. By utilizing the highly interpretable models, a demand-orientated (multi-)optimization coal pyrolysis can be achieved. The bi-objective optimization was conducted on the yield of tar and the content of light aromatics in tar, and the results show that Co is the optimal loading metal.

Unraveling the root causes of faults in mobile communications: A comparative analysis of different model explainability techniques

The escalating demand and complexity of monitoring services handled by Network Operations Centers (NOCs) have led Mobile Network Operators (MNOs) to prioritize automated solutions for network fault detection and diagnosis. Consequently, various Machine Learning (ML)-based Root Cause Analysis (RCA) systems have been developed, however their lack of explainability poses a challenge due to the predominantly black-box nature of ML models. This paper addresses this issue by presenting a supervised clustering methodology capable of integrating both glass-box and black-box models, the latter complemented by post-hoc explainability techniques. While black-box models excel in predictive capabilities, necessitating post-hoc techniques for explainability, glass-box models prioritize transparent decision-making, fostering a clearer understanding of the model’s behavior. This work delineates a methodology for performing RCA of faults in the User Downlink (DL) Average Throughput Key Performance Indicator (KPI), simultaneously comparing the performance of black-box models (Light Gradient-Boosting Machine (LightGBM) and Extreme Gradient Boosting (XGBoost)) with glass-box models (Logistic Regression (LR) and Explainable Boosting Machine (EBM)). Results revealed that the LightGBM black-box algorithm coupled with the SHapley Additive exPlanations (SHAP) method demonstrated superior performance in fault detection and diagnosis, without compromising the overall explainability. Consequently, it was possible to identify faults related to radio conditions, low network usage in specific user groups, low network capacity, and mobility issues. The paper concludes with practical mitigation strategies for each identified fault cluster.

AI-assisted selection of mating pairs through simulation-based optimized progeny allocation strategies in plant breeding.

Emerging technologies such as genomic selection have been applied to modern plant and animal breeding to increase the speed and efficiency of variety release. However, breeding requires decisions regarding parent selection and mating pairs, which significantly impact the ultimate genetic gain of a breeding scheme. The selection of appropriate parents and mating pairs to increase genetic gain while maintaining genetic diversity is still an urgent need that breeders are facing. This study aimed to determine the best progeny allocation strategies by combining future-oriented simulations and numerical black-box optimization for an improved selection of parents and mating pairs. In this study, we focused on optimizing the allocation of progenies, and the breeding process was regarded as a black-box function whose input is a set of parameters related to the progeny allocation strategies and whose output is the ultimate genetic gain of breeding schemes. The allocation of progenies to each mating pair was parameterized according to a softmax function, whose input is a weighted sum of multiple features for the allocation, including expected genetic variance of progenies and selection criteria such as different types of breeding values, to balance genetic gains and genetic diversity optimally. The weighting parameters were then optimized by the black-box optimization algorithm called StoSOO via future-oriented breeding simulations. Simulation studies to evaluate the potential of our novel method revealed that the breeding strategy based on optimized weights attained almost 10% higher genetic gain than that with an equal allocation of progenies to all mating pairs within just four generations. Among the optimized strategies, those considering the expected genetic variance of progenies could maintain the genetic diversity throughout the breeding process, leading to a higher ultimate genetic gain than those without considering it. These results suggest that our novel method can significantly improve the speed and efficiency of variety development through optimized decisions regarding the selection of parents and mating pairs. In addition, by changing simulation settings, our future-oriented optimization framework for progeny allocation strategies can be easily implemented into general breeding schemes, contributing to accelerated plant and animal breeding with high efficiency.

Shaping Up SHAP: Enhancing Stability through Layer-Wise Neighbor Selection

Machine learning techniques, such as deep learning and ensemble methods, are widely used in various domains due to their ability to handle complex real-world tasks. However, their black-box nature has raised multiple concerns about the fairness, trustworthiness, and transparency of computer-assisted decision-making. This has led to the emergence of local post-hoc explainability methods, which offer explanations for individual decisions made by black-box algorithms. Among these methods, Kernel SHAP is widely used due to its model-agnostic nature and its well-founded theoretical framework. Despite these strengths, Kernel SHAP suffers from high instability: different executions of the method with the same inputs can lead to significantly different explanations, which diminishes the relevance of the explanations. The contribution of this paper is two-fold. On the one hand, we show that Kernel SHAP's instability is caused by its stochastic neighbor selection procedure, which we adapt to achieve full stability without compromising explanation fidelity. On the other hand, we show that by restricting the neighbors generation to perturbations of size 1 -- which we call the coalitions of Layer 1 -- we obtain a novel feature-attribution method that is fully stable, computationally efficient, and still meaningful.

A two-stage frequency-domain generation algorithm based on differential evolution for black-box adversarial samples

Adversarial sample generation problem is a hot issue in the security field of deep learning. Evolutionary algorithm has been widely used to solve this problem in recent years because of its good global search ability. However, existing methods still suffer from the “curse of dimensionality” when attacking high-resolution images. In this paper, a two-stage frequency domain generation algorithm of black-box adversarial samples based on differential evolution is proposed. In the first stage, a representative image-guided differential evolution method is proposed to quickly generate a universal adversarial perturbation with a high attack success rate in the frequency-domain. In the second stage, a space reduction strategy based on frequency-domain pixel blocks is designed to reduce the search space and alleviate the problem of “curse of dimensionality”. In addition, a new space–frequency interaction sensitivity measure is introduced to evaluate the similarity between the adversarial samples and the original images. The adversarial perturbations obtained by the measure are more in line with the subjective perception of the human eye. Finally, compared with several typical black-box adversarial sample generation algorithms, experimental results show that the proposed algorithm can achieve higher attack success rate with less prediction times.

PhagoStat a scalable and interpretable end to end framework for efficient quantification of cell phagocytosis in neurodegenerative disease studies

Quantifying the phagocytosis of dynamic, unstained cells is essential for evaluating neurodegenerative diseases. However, measuring rapid cell interactions and distinguishing cells from background make this task very challenging when processing time-lapse phase-contrast video microscopy. In this study, we introduce an end-to-end, scalable, and versatile real-time framework for quantifying and analyzing phagocytic activity. Our proposed pipeline is able to process large data-sets and includes a data quality verification module to counteract potential perturbations such as microscope movements and frame blurring. We also propose an explainable cell segmentation module to improve the interpretability of deep learning methods compared to black-box algorithms. This includes two interpretable deep learning capabilities: visual explanation and model simplification. We demonstrate that interpretability in deep learning is not the opposite of high performance, by additionally providing essential deep learning algorithm optimization insights and solutions. Besides, incorporating interpretable modules results in an efficient architecture design and optimized execution time. We apply this pipeline to quantify and analyze microglial cell phagocytosis in frontotemporal dementia (FTD) and obtain statistically reliable results showing that FTD mutant cells are larger and more aggressive than control cells. The method has been tested and validated on several public benchmarks by generating state-of-the art performances. To stimulate translational approaches and future studies, we release an open-source end-to-end pipeline and a unique microglial cells phagocytosis dataset for immune system characterization in neurodegenerative diseases research. This pipeline and the associated dataset will consistently crystallize future advances in this field, promoting the development of efficient and effective interpretable algorithms dedicated to the critical domain of neurodegenerative diseases’ characterization. https://github.com/ounissimehdi/PhagoStat.

Dynamic Exploration–Exploitation Pareto Approach for high-dimensional expensive black-box optimization

Surrogate Optimization (SO) plays a vital role in optimizing performance parameters for computationally expensive simulations. However, SO encounters significant challenges in high-dimensional spaces due to the curse of dimensionality, hampering effective point sampling around global optima. In this paper, we introduce “Dynamic Exploration–Exploitation Pareto Approach (DEEPA),” a novel SO method that combines Pareto sampling with a dynamic discretization schema to optimize high-dimensional black-box functions. Unlike traditional SO methods that heavily rely on specific surrogate models, Pareto sampling offers a more adaptable approach. Improvement-based acquisition functions, frequently employed in black-box optimization, are sensitive to model accuracy and tend to prioritize exploitation, potentially missing valuable regions of interest in complex landscapes. Furthermore, they can encounter challenges when dealing with high-dimensional problems due to the curse of dimensionality. DEEPA leverages dynamic coordinate importance to generate samples effectively, providing a solution for addressing high-dimensionality and complex functions. We employ feature selection strategies to assign importance probabilities to perturb each coordinate, demonstrating the impact of importance-based perturbation on convergence to a near-optimal region. We showcase DEEPA’s versatility in fixed-batch evaluation environments using complex global optimization test problems with various topological properties. We compare DEEPA’s performance with state-of-the-art black-box optimization algorithms, and our experimental results demonstrate DEEPA’s superior performance, particularly in complex problems with multiple local minima.

Explanation of the influence of geomorphometric variables on the landform classification based on selected areas in Poland

In recent years, automatic image classification methods have significantly progressed, notably black box algorithms such as machine learning and deep learning. Unfortunately, such efforts only focused on improving performance, rather than attempting to explain and interpret how classification models actually operate. This article compares three state-of-the-art algorithms incorporating random forests, gradient boosting and convolutional neural networks for geomorphological mapping. It also attempts to explain how the most effective classifier makes decisions by evaluating which of the geomorphometric variables are most important for automatic mapping and how they affect the classification results using one of the explainable artificial intelligence techniques, namely accumulated local effects (ALE). This method allows us to understand the relationship between predictors and the model’s outcome. For these purposes, eight sheets of the digital geomorphological map of Poland on the scale of 1:100,000 were used as the reference material. The classification results were validated using the holdout method and cross-validation for individual sheets representing different morphogenetic zones. The terrain elevation entropy, absolute elevation, aggregated median elevation and standard deviation of elevation had the greatest impact on the classification results among the 15 geomorphometric variables considered. The ALE analysis was conducted for the XGBoost classifier, which achieved the highest accuracy of 92.8%, ahead of Random Forests at 84% and LightGBM at 73.7% and U-Net at 59.8%. We conclude that automatic classification can support geomorphological mapping only if the geomorphological characteristics in the predicted area are similar to those in the training dataset. The ALE plots allow us to analyze the relationship between geomorphometric variables and landform membership, which helps clarify their role in the classification process.