Search

Quality control is being increasingly automatised in the context of Industry 4.0. Its automatisation reduces inspection times and ensures the same criteria are used to evaluate all products. One of the challenges when developing supervised machine learning models is the availability of labelled data. Few-shot learning promises to be able to learn from few samples and, therefore, reduce the labelling effort. In this work, we combine this approach with unsupervised methods that learn anomaly maps on unlabelled data, providing additional information to the model and enhancing the classification models' discriminative capability. Our results show that the few-shot learning models achieve competitive results compared to those trained in a classical supervised classification setting. Furthermore, we develop novel active learning data sampling strategies to label an initial support set. The results show that using sampling strategies to create and label the initial support set yields better results than selecting samples at random. We performed the experiments on four datasets considering real-world data provided by Philips Consumer Lifestyle BV and Iber-Oleff - Componentes Tecnicos Em Plástico, S.A. [ABSTRACT FROM AUTHOR]

The melting temperature is important for materials design because of its relationship with thermal stability, synthesis, and processing conditions. Current empirical and computational melting point estimation techniques are limited in scope, computational feasibility, or interpretability. We report the development of a machine learning methodology for predicting melting temperatures of binary ionic solid materials. We evaluated different machine-learning models trained on a dataset of the melting points of 476 non-metallic crystalline binary compounds using materials embeddings constructed from elemental properties and density-functional theory calculations as model inputs. A direct supervised-learning approach yields a mean absolute error of around 180 K but suffers from low interpretability. We find that the fidelity of predictions can further be improved by introducing an additional unsupervised-learning step that first classifies the materials before the melting-point regression. Not only does this two-step model exhibit improved accuracy, but the approach also provides a level of interpretability with insights into feature importance and different types of melting that depend on the specific atomic bonding inside a material. Motivated by this finding, we used a symbolic learning approach to find interpretable physical models for the melting temperature, which recovered the best-performing features from both prior models and provided additional interpretability. [ABSTRACT FROM AUTHOR]

For two-dimensional (2D) materials, the exact thickness of the material often dictates its physical and chemical properties. The 2D quantum material WTe2 possesses properties that vary significantly from a single layer to multiple layers, yet it has a complicated crystal structure that makes it difficult to differentiate thicknesses in atomic-resolution images. Furthermore, its air sensitivity and susceptibility to electron beam-induced damage heighten the need for direct ways to determine the thickness and atomic structure without acquiring multiple measurements or transferring samples in ambient atmosphere. Here, we demonstrate a new method to identify the thickness up to ten van der Waals layers in Td-WTe2 using atomic-resolution high-angle annular dark-field scanning transmission electron microscopy image simulation. Our approach is based on analyzing the intensity line profiles of overlapping atomic columns and building a standard neural network model from the line profile features. We observe that it is possible to clearly distinguish between even and odd thicknesses (up to seven layers), without using machine learning, by comparing the deconvoluted peak intensity ratios or the area ratios. The standard neural network model trained on the line profile features allows thicknesses to be distinguished up to ten layers and exhibits an accuracy of up to 94% in the presence of Gaussian and Poisson noise. This method efficiently quantifies thicknesses in Td-WTe2, can be extended to related 2D materials, and provides a pathway to characterize precise atomic structures, including local thickness variations and atomic defects, for few-layer 2D materials with overlapping atomic column positions. [ABSTRACT FROM AUTHOR]

The article presents a response to a study entitled "Competitive Advantages Through Artificial Intelligence: Toward a Theory of Situated AI," published by Ayenda Kemp. The author argues that Kemp's theory regarding the means through which artificial intelligence can promote competitive advantage does not take into account that pursuing AI could likewise have the opposite effect on competitive advantage. The article argues that Kemp's theory does not take into account certain factors, such as the impact AI has on social values, the impact AI has on the nature of the work process itself, and the potential for “algorithmic inertia” whereby the AI may lack the ability to adapt to environmental changes in a dynamic way.

Fault diagnosis is an indispensable basis for the collaborative maintenance in prognostic and health management. Most of existing data-driven fault diagnosis approaches are designed in the framework of supervised learning, which requires a large number of labelled samples. In this paper, a novel self-training semi-supervised deep learning (SSDL) approach is proposed to train a fault diagnosis model together with few labelled and abundant unlabelled samples. The addressed SSDL approach is realised by initialising a stacked sparse auto-encoder classifier using the labelled samples, and subsequently updating the classifier via sampling a few candidates with most reliable pseudo labels from the unlabelled samples step by step. Unlike the commonly used static sampling strategy in existing self-training semi-supervised frameworks, a gradually exploiting mechanism is proposed in SSDL to increase the number of selected pseudo-labelled candidates gradually. In addition, instead of using the prediction accuracy as the confidence estimation for pseudo-labels, a distance-based sampling criterion is designed to assign the label for each unlabelled sample by its nearest labelled sample based on their Euclidean distances in the deep feature space. The experimental results show that the proposed SSDL approach can achieve good prediction accuracy compared to other self-training semi-supervised learning algorithms. [ABSTRACT FROM AUTHOR]

The thermophysical properties of liquid metals and alloys are crucial to explore the intrinsic mechanisms of the solidification process, glass formation, and fluid dynamics. The deep learning approaches have emerged as powerful tools in numerous scientific fields and exhibit extraordinary accuracy in the estimation of physical properties and structural characteristics for various materials. In this Tutorial, focusing on the thermophysical properties of liquid multicomponent alloys, deep learning methods, including both supervised learning and active learning, are introduced. Combined with the verification from electrostatic and electromagnetic levitation experiments, the influences of training parameters and methods on the accuracy to obtain interatomic potential by deep learning are revealed on the basis of deep neural network algorithm. As a result, this prediction method of liquid state properties for multicomponent alloys exhibited the dual advantages of high accuracy derived from density functional theory and low computational cost associated with empirical potential. [ABSTRACT FROM AUTHOR]

This study addresses the potential for improving the accuracy, scope, and value of sales research through the application of data mining and machine learning algorithms. By examining prior research, identifying opportunities for improvement, and assessing gaps that can benefit from machine learning, research and application are made more accessible for sales researchers and managers. Machine learning can address important sales research questions that cannot be answered with the same accuracy or efficiency as traditional research methods. This study demonstrates the benefits of the methods through an example of application to the prediction of salesforce performance based on behavioral, attitudinal, and demographic data. This includes future research ideas, usage cases, and applications where machine learning could advance sales research and management. Machine learning and predictive analytics methods have multiple applications, including in B2C and B2B market contexts and for companies and independent sales teams. [ABSTRACT FROM AUTHOR]

This work uses supervised learning to optimize the package design with validated experimental results for piezoelectric micromachined ultrasonic transducers (PMUTs) to increase and alter the sound pressure level (SPL). Advancements as compared to the state-of-art include: (1) a neural network model to achieve a mean squared error of less than 0.65 dB2 post 100 epochs; (2) increased vibration amplitude by 17.9 dBV at the first-mode resonance frequency of 33.5 kHz; and (3) SPL enhancements below the 20 kHz frequency range such as the magnitude increases of more than 60 dBV at 5 kHz. As such, the package design shifts the emitting acoustic energy from the ultrasound to audio range in favor of various applications, including audio speakers.

Emoji are commonly used in social media to convey affects, emotions, and attitudes. While popular in social media, their use in educational contexts has been sparsely studied even though emoji can be a natural way for students to express what they are feeling about the learning material. This paper studies how students use instructor-selected emoji when relating to and engaging with educational content. We use an online platform for collaborative annotations where discussions are embedded into the readings for the course. We also make it possible for students to use 11 unique emoji-hashtag pairings to express their thoughts and feelings about the readings and the ongoing discussion. We provide an empirical analysis of the usage of these emoji-hashtag pairs by over 1,800 students enrolled in different offerings of an introductory biology course from multiple academic terms. We also introduce a heat map, which allows the instructional team to visualize the distribution and types of emoji used by students in different parts of the reading material. To evaluate the heat map, we conducted a user study with five instructors/TAs. We found that instructors/TAs use the heat map as a tool for identifying textbook sections that students find difficult and/or interesting and plan to use it to help them design the online content for future classes. Finally, we introduce a computational analysis for predicting emoji/hashtag pairs based on the content of a given student post. We use pre-trained deep learning language models (BERT) to predict the emoji attached to a student’s post and then study the extent to which this model generated in an introductory biology course can be generalized to predict student emoji usage in other courses.

One of the most prevalent psychiatric disorders is major depressive disorder (MDD), which increases the probability of suicidal ideation or untimely demise. Abnormal frontal hemodynamic changes detected by functional near-infrared spectroscopy (fNIRS) during verbal fluency task (VFT) have the potential to be used as an objective indicator for assessing clinical symptoms. However, comprehensive quantitative and objective assessment instruments for individuals who exhibit symptoms suggestive of depression remain undeveloped. Drawing from a total of 467 samples in a large-scale dataset comprising 289 MDD patients and 178 healthy controls, fNIRS measurements were obtained throughout the VFT. To identify unique MDD biomarkers, this research introduced a data representation approach for extracting spatiotemporal features from fNIRS signals, which were subsequently utilized as potential predictors. Machine learning classifiers (e.g., Gradient Boosted Decision Trees (GBDT) and Multilayer Perceptron) were implemented to assess the ability to predict selected features. The mean and standard deviation of the cross-validation indicated that the GBDT model, when combined with the 180-feature pattern, distinguishes patients with MDD from healthy controls in the most effective manner. The accuracy of correct classification for the test set was 0.829 ± 0.053, with an AUC of 0.895 (95 % CI: 0.864–0.925) and a sensitivity of 0.914 ± 0.051. Channels that made the most important contribution to the identification of MDD were identified using Shapley Additive Explanations method, located in the frontopolar area and the dorsolateral prefrontal cortex, as well as pars triangularis Broca's area. Assessment of abnormal prefrontal activity during the VFT in MDD serves as an objectively measurable biomarker that could be utilized to evaluate cognitive deficits and facilitate early screening for MDD. The model suggested in this research could be applied to large-scale case-control fNIRS datasets to detect unique characteristics of MDD and offer clinicians an objective biomarker-based analytical instrument to assist in the evaluation of suspicious cases. • fNIRS during VFT has the potential to be used as an objective indicator for assessing MDD. • The study developed an interpretable machine learning framework for screening patients with MDD from healthy controls. • The classification of model for MDD diagnosis had a mean accuracy of 0.829, with an AUC of 0.895. • The model could be applied to large-scale case-control fNIRS datasets to identify neuroimaging biomarkers and offer clinicians an objective instrument. [ABSTRACT FROM AUTHOR]

The AUTO algorithm is presented to incrementally build models and solve supervised learning problems. AUTO uses traditional derivative-free optimisation, like genetic algorithms, to search the problem space of arbitrary functions. With reinforcement learning, AUTO learns actions to guide its search process and gradually improve performance. A comparative analysis is presented exploring supervised learning performance and interpretability of AUTO models. Results indicate AUTO outperforms genetic programming and rivals multilayer perceptrons. AUTO automatically builds models to closely match datasets without the limited search space of traditional supervised learning. With enough time and computational resources, AUTO can generate any function. [ABSTRACT FROM AUTHOR]

OBJECTIVE: Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG) are effective procedures to treat and manage type 2 diabetes (T2D). However, the underlying metabolic adaptations that mediate improvements in glucose homeostasis remain largely elusive. The purpose of this study was to identify metabolic signatures associated with biochemical resolution of T2D after medical therapy (MT) or bariatric surgery. RESEARCH DESIGN AND METHODS: Plasma samples from 90 patients (age 49.9 ± 7.6 years; 57.7% female) randomly assigned to MT (n = 30), RYGB (n = 30), or SG (n = 30) were retrospectively subjected to untargeted metabolomic analysis using ultra performance liquid chromatography with tandem mass spectrometry at baseline and 24 months of treatment. Phenotypic importance was determined by supervised machine learning. Associations between change in glucose homeostasis and circulating metabolites were assessed using a linear mixed effects model. RESULTS: The circulating metabolome was dramatically remodeled after SG and RYGB, with largely overlapping signatures after MT. Compared with MT, SG and RYGB profoundly enhanced the concentration of metabolites associated with lipid and amino acid signaling, while limiting xenobiotic metabolites, a function of decreased medication use. Random forest analysis revealed 2-hydroxydecanoate as having selective importance to RYGB and as the most distinguishing feature between MT, SG, and RYGB. To this end, change in 2-hydroxydecanoate correlated with reductions in fasting glucose after RYGB but not SG or MT. CONCLUSIONS: We identified a novel metabolomic fingerprint characterizing the longer-term adaptations to MT, RYGB, and SG. Notably, the metabolomic profiles of RYGB and SG procedures were distinct, indicating equivalent weight loss may be achieved by divergent effects on metabolism. [ABSTRACT FROM AUTHOR]

Carbon fibre reinforced polymer (CFRP) composites are widely used in the aerospace industry; they are susceptible to impact damage from low-velocity impacts throughout their service life. Long pulse thermography (LPT) was proposed to detect defect in low-velocity impact test (LVIT). In this framework, the defects of matrix fractures, fibre cracks and interface delamination were produced by LVIT in the CFRP specimens. The halogen lamps LPT test system was applied to detect the low-velocity impact damage, and an A655sc infrared camera was applied to collect the surface temperature. In order to improve the efficiency of the impact damage detection, the infrared image sequences were processed by applying different post-processing techniques such as pulse phase thermography (PPT), thermal signal reconstruction (TSR), principal component analysis (PCA), total harmonic distortion (THD), linear discriminant analysis (LDA) and quadratic discriminant analysis (QDA). The results show that the signal-to-noise ratio (SNR) was improved by PPT, TSR, PCA, THD, LDA and QDA post-processing techniques. The proposed supervised learning (SL) QDA algorithm achieves better defect identification results than the above post-processing techniques. The detectability of real defects can provide a realistic assessment by the above post-processing techniques. [ABSTRACT FROM AUTHOR]

Similar waveforms and overlapping frequency ranges make distinguishing blasting vibrations and microseismic signals challenging, causing interference with coal mine microseismic monitoring systems. To address this problem, we propose a spectrum dataset (MSData) reflecting the spectrum features of both signal types and present a signal classification network (SCNet) combining CNNs and Transformers for signal classification. The network can learn multi-dimensional features of both signals from MSData and automatically and efficiently identify the two signal types. Experimental results yield F1-scores of 0.991 for microseismic signals and 0.993 for blasting vibration signals, meeting engineering application requirements. [ABSTRACT FROM AUTHOR]

Transformer-based models for transfer learning have the potential to achieve high prediction accuracies on text-based supervised learning tasks with relatively few training data instances. These models are thus likely to benefit social scientists that seek to have as accurate as possible text-based measures, but only have limited resources for annotating training data. To enable social scientists to leverage these potential benefits for their research, this article explains how these methods work, why they might be advantageous, and what their limitations are. Additionally, three Transformer-based models for transfer learning, BERT, RoBERTa, and the Longformer, are compared to conventional machine learning algorithms on three applications. Across all evaluated tasks, textual styles, and training data set sizes, the conventional models are consistently outperformed by transfer learning with Transformers, thereby demonstrating the benefits these models can bring to text-based social science research. [ABSTRACT FROM AUTHOR]

Over 40,000 people die by suicide annually in the U.S., and firearms are the most lethal suicide method. There is limited evidence on the effectiveness of many state-level policies on reducing firearm suicide. The objective of this study was to identify public policies that best predict state-level firearm suicide rates. Data from the Centers for Disease Control and Prevention's WONDER system and the State Firearm Law Database, a longitudinal catalog of 134 firearm safety laws, were analyzed. The analysis included 1,450 observations from 50 states spanning 1991–2019. An ElasticNet regression technique was used to analyze the relationship between the policy variables and firearm suicide rates. Nested cross-validation was performed to tune the model hyperparameters. The study data were collected and analyzed in 2023 and 2024. The optimized ElasticNet approach had a mean squared error of 2.07, which was superior to nonregularized and dummy regressor models. The most influential policies for predicting the firearm suicide rate on average included laws requiring firearm dealers that sell handguns to have a state license and laws requiring individuals to obtain a permit to purchase a firearm through an approval process that includes law enforcement, among others. On average, firearm suicide rates were lower in state-years that had each influential policy active. Notably, these analyses were ecological and noncausal. However, this study was able to use a supervised machine learning approach with inherent feature selection and many policy types to make predictions using unseen data (i.e., balancing Lasso and Ridge regularization penalties). [ABSTRACT FROM AUTHOR]

Athletes' performance evaluation is a critical step for the assessment of the skills and the quality of training of the athletes. This is even more important in team sports such as football, in which different roles, and hence different skills, are required. In light of this, the presented work aims at forecasting above team-average performance of football players by using supervised machine learning algorithms. Such algorithms were trained and tested on four biometric parameters as input and seven performance indicators as labels. The algorithms have been optimized using three optimization techniques: the gridsearch, and two versions of the whale-optimization algorithm, the standard one and another, proposed by us, in which Euclidean distance is used. The analyses were conducted by dividing the players by their role: the strikers, the midfielders and the defenders, to take into account the different skills required for each task. The obtained results show that the Random Forest, trained with specific combinations of biometric parameters is capable of predicting the selected performance indicators with an accuracy higher than 90%, for all the considered players' roles and that the proposed optimization technique outperforms the existing ones. In light of these results, the proposed method could be used profitably by football teams' staff to monitor their players, offering the possibility to take tactical decisions, design customized training sessions and orient market choices. [ABSTRACT FROM AUTHOR]

Controlled parking systems in cities provide designated parking zones and allow citizens to easily find parking spaces increasing comfort and potentially reducing traffic and pollution. However, illegally occupied parking spaces can negatively affect the accuracy of these systems, diminishing their benefits. This work presents a semi-supervised deep learning approach for fine-grained parking violation rate prediction using graph neural networks, which can capture both the spatial and temporal dynamics of parking systems and driver behavior, providing a valuable tool for controlled parking systems. The proposed method addresses several challenges faced in this task, including the need to appropriately construct temporal graph-based training datasets and design a model that can handle missing values. Additionally, the method includes a novel semi-supervised data augmentation and smoothing technique to mitigate the effect of the small amount of annotated data and/or missing data in the training dataset. The proposed method was evaluated using a large-scale dataset from Thessaloniki's on-street public parking system and found to significantly improve the accuracy of parking violation prediction compared to state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

Failure mode identification in unreinforced masonry (URM) walls is a challenging task because of the presence of multiple damage mechanisms, including so-called hybrid modes. This paper employed a novel application of supervised and unsupervised learning to link URM wall design and mechanical properties to the failure mode. We used a database with 330 backbone curves from cyclically loaded URM walls as well as the associated images captured during the experiments. Based on the observations documented during the experiments, information on the cyclic curves, and the associated images, the walls were first manually classified into four failure modes: bed-joint sliding, diagonal tension, rocking, and toe crushing. Then k -means clustering was used to group the damaged walls into four classes based on critical points along the backbone curves. A hybridity index was introduced to quantify the contribution of each failure mode based on the distance from the centroid of the clusters. The design and mechanical properties of the URM walls were then used to predict the hybridity index using a multioutput regression model. The hybridity prediction model determines the contribution of the various failure modes to the ultimate behavior of a damaged URM wall. The proposed framework provides a robust approach to quantifying the relative contribution of each failure mechanism to the overall performance of the URM wall. [ABSTRACT FROM AUTHOR]

Nuclear forensics relies on different signatures to identify the source of nuclear material. Such signatures include crystalline structure, chemical composition, and particle morphology. One way to quantify morphology in electron microscope imagery is through image segmentation, where pixels are assigned to several partitions (or groups) that correspond to particles, grains, and other objects of interest within the image. Once pixels are assigned to segments, it is relatively straightforward to quantify other quantities of interest, such as grain size, circularity, etc. However, the range and diversity of microscope images make it difficult to obtain an accurate segmentation automatically. The accuracy of segmentation can be improved through supervised learning, but this requires many images to be manually segmented. Another way to improve the accuracy is to use interactive segmentation. Interactive segmentation requires a human to provide image-specific user input to improve performance. However, the amount of user input (effort) is generally far less than is required for supervised learning. In this paper, we investigate several parallelization strategies to automatically explore the user input parameter space of interactive segmentation algorithms across a large number of images. Specifically, we investigate four different parallelization mechanisms in a high-performance computing (HPC) environment and use the Amdahl fraction to evaluate efficiency on multiple processor cores across multiple nodes. Ultimately, the parallelization strategy that was most efficient utilized the message passing interface integrated with the segmentation and quantification code. This strategy had an Amdahl fraction of 0.985 and a performance of about 0.251 s/image. These results indicate that the parameter space of interactive segmentation algorithms can be efficiently explored using HPC. This opens the door to future work where user input is reduced and where interactive image segmentation algorithms are automatically applied to large image sets. [ABSTRACT FROM AUTHOR]

Engineering materials in constructed systems in service exhibit complex patterns, including structural damage, environmental artifacts, and artificial anomalies. In recent years, machine vision methods have been extensively studied, most of which train models using regular grey or color images in the visible bands and label at pixel levels with a large volume of data. The authors propose using hyperspectral imaging (HSI) for structural material condition assessment in this work. Compared with visible images, the research challenge is that HSI pixels with high-dimensional spectral profiles are beyond human perceptive capabilities with hidden discriminative power. Learning from labeled and unlabeled data is one direct approach to unlocking this power. A deep neural network-enabled spatial-spectral feature extraction and a semisupervised learning architecture were developed in this work. A human-in-the-loop (HITL) framework was comparatively studied with three incremental training-data configuration schemes. The paper concludes with the following empirical findings: (1) fully supervised learning determines the baseline of the detection performance; (2) an extensive range of ratio values exists between the unlabeled and the labeled data for incremental semisupervised learning, and a 1∶1 ratio can be taken as a conservative and operational ratio; and (3) with parametric semisupervised learning with equal labeled and unlabeled data participation, the proposed HITL operational workflow can be implemented as a practical approach for HSI-based structural material and damage detection. [ABSTRACT FROM AUTHOR]

• Predictive models using immune-based biomarkers perform well in discerning bipolar and schizophrenic patients from controls. • IL-6, TNF-ɑ, IFN-γ QUINO and PICO are identified as key markers for schizophrenia and bipolar disorder classification. • Challenges in differentiating schizophrenia from bipolar disorder highlight shared immunological mechanisms. • Larger multi-centric studies and multi-domain models are needed for enhanced generalizability and clinical translation. Schizophrenia and bipolar disorder frequently face significant delay in diagnosis, leading to being missed or misdiagnosed in early stages. Both disorders have also been associated with trait and state immune abnormalities. Recent machine learning-based studies have shown encouraging results using diagnostic biomarkers in predictive models, but few have focused on immune-based markers. Our main objective was to develop supervised machine learning models to predict diagnosis and illness state in schizophrenia and bipolar disorder using only a panel of peripheral kynurenine metabolites and cytokines. The cross-sectional I-GIVE cohort included hospitalized acute bipolar patients (n = 205), stable bipolar outpatients (n = 116), hospitalized acute schizophrenia patients (n = 111), stable schizophrenia outpatients (n = 75) and healthy controls (n = 185). Serum kynurenine metabolites, namely tryptophan (TRP), kynurenine (KYN), kynurenic acid (KA), quinaldic acid (QUINA), xanthurenic acid (XA), quinolinic acid (QUINO) and picolinic acid (PICO) were quantified using liquid chromatography-tandem mass spectrometry (LC-MS/MS), while V-plex Human Cytokine Assays were used to measure cytokines (interleukin-6 (IL-6), IL-8, IL-17, IL-12/IL23-P40, tumor necrosis factor-alpha (TNF-ɑ), interferon-gamma (IFN-γ)). Supervised machine learning models were performed using JMP Pro 17.0.0. We compared a primary analysis using nested cross-validation to a split set as sensitivity analysis. Post-hoc, we re-ran the models using only the significant features to obtain the key markers. The models yielded a good Area Under the Curve (AUC) (0.804, Positive Prediction Value (PPV) = 86.95; Negative Prediction Value (NPV) = 54.61) for distinguishing all patients from controls. This implies that a positive test is highly accurate in identifying the patients, but a negative test is inconclusive. Both schizophrenia patients and bipolar patients could each be separated from controls with a good accuracy (SCZ AUC 0.824; BD AUC 0.802). Overall, increased levels of IL-6, TNF-ɑ and PICO and decreased levels of IFN-γ and QUINO were predictive for an individual being classified as a patient. Classification of acute versus stable patients reached a fair AUC of 0.713. The differentiation between schizophrenia and bipolar disorder yielded a poor AUC of 0.627. This study highlights the potential of using immune-based measures to build predictive classification models in schizophrenia and bipolar disorder, with IL-6, TNF-ɑ, IFN-γ, QUINO and PICO as key candidates. While machine learning models successfully distinguished schizophrenia and bipolar disorder from controls, the challenges in differentiating schizophrenic from bipolar patients likely reflect shared immunological pathways by the both disorders and confounding by a larger state-specific effect. Larger multi-centric studies and multi-domain models are needed to enhance reliability and translation into clinic. [ABSTRACT FROM AUTHOR]

Group recommendation involves suggesting items or activities to a group of users based on their collective preferences or characteristics. Graph contrastive learning is a technique used to learn representations of items and users in a graph structure. Although contrastive learning-based recommendation techniques reduce the data sparsity problem by extracting general features from raw data and also make the representation of user-item bipartite graph augmentations more consistent, the factors contributing to improving the performance of this technique are still not fully understood. Meanwhile, graph augmentations have little importance in contrastive learning-based recommendation and are relatively unreliable. The eXtremely Simple Graph Contrastive Learning (XSimGCL) provides novel insights into the effect of contrastive learning on recommendation, where views for contrastive learning are created through a simple yet effective noise-based embedding augmentation. Although XSimGCL infers the final group decision by dynamically aggregating the preferences of group members and includes various types of interaction, the performance of supervised learning is reduced due to the data sparsity problem, and as a result, the efficiency of group preference representation is limited. To address this challenge, we developed a Group Recommendation model based on XsimGCL in this study (GR-GCL). GR-GCL is inspired by the Light Graph Convolution Network (LightGCN) to realize simultaneous learning of multiple graphs, where initial embedding is considered the only update parameter. Also, GR-GCL improves group recommendation by applying cross-layer contrastive learning in the XSimGCL model by representing more diverse entities. The rationality analysis of our proposed GR-GCL has been performed on several datasets from both analytical and empirical perspectives. Although our model is very simple, it performs better in group recommendations by adjusting the uniformity of representations learned from counterparts based on contrastive learning. [ABSTRACT FROM AUTHOR]

Machine learning algorithms are commonly used for quickly and efficiently counting people from a crowd. Test-time adaptation methods for crowd counting adjust model parameters and employ additional data augmentation to better adapt the model to the specific conditions encountered during testing. The majority of current studies concentrate on unsupervised domain adaptation. These approaches commonly perform hundreds of epochs of training iterations, requiring a sizable number of unannotated data of every new target domain apart from annotated data of the source domain. Unlike these methods, we propose a meta-test-time adaptive crowd counting approach called CrowdTTA, which integrates the concept of test-time adaptation into the meta-learning framework and makes it easier for the counting model to adapt to the unknown test distributions. To facilitate the reliable supervision signal at the pixel level, we introduce uncertainty by inserting the dropout layer into the counting model. The uncertainty is then used to generate valuable pseudo labels, serving as effective supervisory signals for adapting the model. In the context of meta-learning, one image can be regarded as one task for crowd counting. In each iteration, our approach is a dual-level optimization process. In the inner update, we employ a self-supervised consistency loss function to optimize the model so as to simulate the parameters update process that occurs during the test phase. In the outer update, we authentically update the parameters based on the image with ground truth, improving the model's performance and making the pseudo labels more accurate in the next iteration. At test time, the input image is used for adapting the model before testing the image. In comparison to various supervised learning and domain adaptation methods, our results via extensive experiments on diverse datasets showcase the general adaptive capability of our approach across datasets with varying crowd densities and scales. [ABSTRACT FROM AUTHOR]

The surgery-first approach (SFA) orthognathic surgery can be beneficial due to reduced overall treatment time and earlier profile improvement. The objective of this study was to utilize deep learning to predict the treatment modality of SFA or the orthodontics-first approach (OFA) in orthognathic surgery patients and assess its clinical accuracy. A supervised deep learning model using three convolutional neural networks (CNNs) was trained based on lateral cephalograms and occlusal views of 3D dental model scans from 228 skeletal Class III malocclusion patients (114 treated by SFA and 114 by OFA). An ablation study of five groups (lateral cephalogram only, mandible image only, maxilla image only, maxilla and mandible images, and all data combined) was conducted to assess the influence of each input type. The results showed the average validation accuracy, precision, recall, F1 score, and AUROC for the five folds were 0.978, 0.980, 0.980, 0.980, and 0.998 ; the average testing results for the five folds were 0.906, 0.986, 0.828, 0.892, and 0.952. The lateral cephalogram only group had the least accuracy, while the maxilla image only group had the best accuracy. Deep learning provides a novel method for an accelerated workflow, automated assisted decision-making, and personalized treatment planning. [ABSTRACT FROM AUTHOR]

Bipolar disorder has become one of the major mental health issues due to stressed life around the world. This is the major reason for suicides these days as these people are unable to convey their feeling and emotions to others. This proposed research shows the logistic regression and support vector machine hybrid model to predict bipolar disorder in patients is to develop an accurate and reliable model that can effectively predict the presence of bipolar disorder in patients based on their clinical and demographic data. The purpose is to make a framework that can help healthcare professionals diagnose bipolar disorder early, thereby enabling timely and appropriate treatment to be provided. The model should take into account various patient-specific features, such as age, gender, family history, medication use, and other medical conditions, in addition to relevant clinical and demographic variables. The aim is to create a model that can accurately classify patients with bipolar disorder and non-bipolar disorder patients while minimizing false-positive and false-negative classifications. The work shows improvement in evaluation detection in performance with hybrid logistic support vector regression (LSVR) to detect disorder and protect them to avoid worse situation. [ABSTRACT FROM AUTHOR]

Automatic surgical workflow analysis, which aims to recognize the ongoing surgical events in videos, is fundamental for developing context-aware computer-assisted systems. This paper reviews representative surgical workflow recognition algorithms based on deep learning, outlining their merits, limitations, and future research directions. The literature survey was performed on three large bibliographic databases, covering 67 lary sources, which were comparatively analyzed in terms of spatial feature modeling, spatio-temporal feature modeling, input pre-processing, regularization and post-processing algorithms, as well as learning strategies. Then, common public datasets and evaluation metrics for surgical workflow recognition are also described in detail. Finally, we discuss all literature from different perspectives, and point out the challenges, possible solutions and future trends. The need for more diverse and larger datasets, the potential of unsupervised and semi-supervised learning approaches, comprehensive and equitable metrics, establishing complete regulatory and data standards, and interoperability will be key challenges in translating models to clinical operating rooms. And we propose that surgical activity anticipation and employing large language model as training assistant are interesting research directions in surgical workflow analysis. [ABSTRACT FROM AUTHOR]

Purpose: Multiparametric arterial spin labeling (MP‐ASL) can quantify cerebral blood flow (CBF) and arterial cerebral blood volume (CBVa). However, its accuracy is compromised owing to its intrinsically low SNR, necessitating complex and time‐consuming parameter estimation. Deep neural networks (DNNs) offer a solution to these limitations. Therefore, we aimed to develop simulation‐based DNNs for MP‐ASL and compared the performance of a supervised DNN (DNNSup), physics‐informed unsupervised DNN (DNNUns), and the conventional lookup table method (LUT) using simulation and in vivo data. Methods: MP‐ASL was performed twice during resting state and once during the breath‐holding task. First, the accuracy and noise immunity were evaluated in the first resting state. Second, CBF and CBVa values were statistically compared between the first resting state and the breath‐holding task using the Wilcoxon signed‐rank test and Cliff's delta. Finally, reproducibility of the two resting states was assessed. Results: Simulation and first resting‐state analyses demonstrated that DNNSup had higher accuracy, noise immunity, and a six‐fold faster computation time than LUT. Furthermore, all methods detected task‐induced CBF and CBVa elevations, with the effect size being larger with the DNNSup (CBF, p = 0.055, Δ = 0.286; CBVa, p = 0.008, Δ = 0.964) and DNNUns (CBF, p = 0.039, Δ = 0.286; CBVa, p = 0.008, Δ = 1.000) than that with LUT (CBF, p = 0.109, Δ = 0.214; CBVa, p = 0.008, Δ = 0.929). Moreover, all the methods exhibited comparable and satisfactory reproducibility. Conclusion: DNNSup outperforms DNNUns and LUT with respect to estimation performance and computation time. [ABSTRACT FROM AUTHOR]

Direct imaging of exoplanets is particularly challenging due to the high contrast between the planet and the star luminosities, and their small angular separation. In addition to tailored instrumental facilities implementing adaptive optics and coronagraphy, post-processing methods combining several images recorded in pupil tracking mode are needed to attenuate the nuisances corrupting the signals of interest. Most of these post-processing methods build a model of the nuisances from the target observations themselves, resulting in strongly limited detection sensitivity at short angular separations due to the lack of angular diversity. To address this issue, we propose to build the nuisance model from an archive of multiple observations by leveraging supervised deep learning techniques. The proposed approach casts the detection problem as a reconstruction task and captures the structure of the nuisance from two complementary representations of the data. Unlike methods inspired by reference differential imaging, the proposed model is highly non-linear and does not resort to explicit image-to-image similarity measurements and subtractions. The proposed approach also encompasses statistical modelling of learnable spatial features. The latter is beneficial to improve both the detection sensitivity and the robustness against heterogeneous data. We apply the proposed algorithm to several data sets from the VLT/SPHERE instrument, and demonstrate a superior precision-recall trade-off compared to the PACO algorithm. Interestingly, the gain is especially important when the diversity induced by ADI is the most limited, thus supporting the ability of the proposed approach to learn information across multiple observations. [ABSTRACT FROM AUTHOR]

Neural processes (NPs) are models for meta-learning which output uncertainty estimates. So far, most studies of NPs have focused on low-dimensional datasets of highly-correlated tasks. While these homogeneous datasets are useful for benchmarking, they may not be representative of realistic transfer learning. In particular, applications in scientific research may prove especially challenging due to the potential novelty of meta-testing tasks. Molecular property prediction is one such research area that is characterized by sparse datasets of many functions on a shared molecular space. In this paper, we study the application of graph NPs to molecular property prediction with DOCKSTRING, a diverse dataset of docking scores. Graph NPs show competitive performance in few-shot learning tasks relative to supervised learning baselines common in chemoinformatics, as well as alternative techniques for transfer learning and meta-learning. In order to increase meta-generalization to divergent test functions, we propose fine-tuning strategies that adapt the parameters of NPs. We find that adaptation can substantially increase NPs' regression performance while maintaining good calibration of uncertainty estimates. Finally, we present a Bayesian optimization experiment which showcases the potential advantages of NPs over Gaussian processes in iterative screening. Overall, our results suggest that NPs on molecular graphs hold great potential for molecular property prediction in the low-data setting. Scientific contribution: Neural processes are a family of meta-learning algorithms which deal with data scarcity by transferring information across tasks and making probabilistic predictions. We evaluate their performance on regression and optimization molecular tasks using docking scores, finding them to outperform classical single-task and transfer-learning models. We examine the issue of generalization to divergent test tasks, which is a general concern of meta-learning algorithms in science, and propose strategies to alleviate it. [ABSTRACT FROM AUTHOR]

The present study assesses the potential of using machine learning (ML) methods to predict the extent of lightning thermal damage in fiber-reinforced composite laminates using three supervised machine learning (SML) algorithms: (1) linear regression (LR), (2) decision tree (DT)-based, and (3) MLP models. These models were based on the 10 most significant factors that influence the severity of lightning damage, including three current waveform parameters, four material configurations, and three orthogonal electrical conductivities of each composite. All models demonstrated good performance with coefficient of determination (R2) values between 0.84 ~ 0.96. The multilayer perceptron (MLP) regression model trained with the lightning matrix damage dataset showed the most promising results (R2 > 0.94). Additional hyperparameter optimization was performed to improve the prediction performance of the baseline MLP model. The hyperparameter optimization (Adam optimizer, tanh activation function, and three hidden layers with 234 neurons) slightly improved the performance of the baseline MLP model by ~0.02, but achieved faster convergence. This result suggests that the baseline MLP model trained with the lightning matrix damage dataset is sufficiently accurate and robust. This paper highlights that ML-informed regression models can serve as an efficient first pass-estimator of lightning matrix damage in composite laminates, potentially reducing the amount of extremely time-consuming and expensive laboratory-scale lightning tests or streamlining the development of complex lightning damage models for future design. [ABSTRACT FROM AUTHOR]

Despite numerous screening tools for colorectal cancer (CRC), 25 % of patients are diagnosed with advanced disease. Novel diagnostic technologies that are early, accurate, and rapid are imperative to assess the therapeutic efficacy of clinical drugs and identify new biomarkers of treatment response. Here Raman spectroscopy (RS) was used to track metabolic reprogramming in KRAS‐mutant HCT116 and SW837 cells, and KRAS wild‐type CC cells. RS combined with multivariate analysis methods distinguished nonresponsive, partially responsive, and responsive cells treated with cetuximab, a monoclonal antibody for EGFR inhibition, sotorasib, a clinically approved KRAS inhibitor, and various doses of trametinib, an inhibitor of the MAPK pathway. Cells treated with a combination of subtoxic doses of trametinib and BKM120, an inhibitor of the PI3K pathway, showed a synergistic response between the two pathways. Using a supervised machine learning regression model, we established a scoring methodology trained to a priori predict therapeutic response to new treatment combinations. RS metabolites were verified with mass spectrometry, and enrichment pathways were identified, including amino acid, purine, and nicotinate and nicotinamide metabolism that differentiated monotherapy from combination therapy. Our approach may ultimately be applicable to patient‐derived primary cells and cultures of patient tumors to predict effective drugs for individualized care. [ABSTRACT FROM AUTHOR]

Polycystic Ovary Syndrome (PCOS) is a recurring endocrine disorder that primarily affects women of reproductive age. It is difficult to diagnose due to its heterogeneous characteristics and overlapping symptoms with other illnesses. As a result, accurate and trustworthy prediction models are required to detect PCOS early. This research work aims to develop ML methods that predict the risk of PCOS among women based on demographic and clinical features. The entire framework is divided into four phases: in Phase I of the study, the SMOTE-Tomek Links (SMTL) technique balances the data set by combining oversampling and undersampling approaches. A novel meta-heuristic-based feature selection approach, the Grey Wolf Optimization (GWO) method, has been employed to select the most crucial features from the dataset, explained in Phase II. Subsequently, in Phase III, a hybridized classifier PODBoost algorithm (Polycystic Ovarian Disorder Boosting algorithm) is devised for faithful early prediction of PCOS using the concepts of different classical supervised learning algorithms. Finally, Explainable AI (XAI) such as the Local Interpretable Model-Agnostic Explanations (LIME) tool has been implemented to interpret relevant predictions made by the proposed classifier. The proposed algorithm is examined utilizing numerous metrics such as Accuracy, Error-Rate, ROC-AUC Score, Recall, Precision, and F1-Score. Among all the evaluated models, the proposed hybridized model has shown an impressive performance with an exceptional accuracy of 97.42 % , indicating its superiority by delivering outstanding results. Based on the findings, the novel meta-heuristic-based feature selection method significantly impacts the outcomes of the proposed hybridized PODBoost algorithm. This algorithm may be recommended to predict PCOS or other relevant diseases having datasets which are multimodal in nature. [ABSTRACT FROM AUTHOR]

Financial technology is crucial for the sustainable development of financial systems. Algorithmic trading, a key area in financial technology, involves automated trading based on predefined rules. However, investors cannot manually analyze all market patterns and establish rules, necessitating the development of supervised learning trading systems that can discover market patterns using machine or deep learning techniques. Many studies on supervised learning trading systems rely on up–down labeling based on price differences, which overlooks the issues of nonstationarity, complexity, and noise in stock data. Therefore, this study proposes an N‐period volatility trading system that addresses the limitations of up–down labeling systems. The N‐period volatility trading system measures price volatility to address uncertainty and enables the construction of a stable, long‐term trading system. Additionally, an instance‐selection technique is utilized to address the limitations of stock data, including noise, nonlinearity, and complexity, while effectively reducing the data size. The effectiveness of the proposed model is evaluated through trading simulations of stocks comprising the NASDAQ 100 index and compared with up–down labeling trading systems. The experimental results demonstrate that the proposed N‐period volatility trading system exhibits higher stability and profitability than other trading systems. [ABSTRACT FROM AUTHOR]

Due to the potentially high cost and difficulty of obtaining labeled data in medical image classification, semi-supervised learning can effectively utilize limited labeled data and large amounts of unlabeled data to enhance the performance of classification models. However, most existing methods utilize pseudo-labeling techniques that filter out a large amount of data, which makes the unlabeled data underutilized. Moreover, the consistency regularization methods rely heavily on the perturbation function suffering from a lack of generalization ability. In this paper, we propose a semi-supervised medical image classification approach based on class prototype matching for soft pseudo-labels with consistency regularization. A class prototype matching module is used to predict soft pseudo-labels for unlabeled data, and the quality of the predicted pseudo-labels is improved by a meticulously designed cache queue with dynamic and unbiased updates. The linear mixture strategy is employed to handle both labeled and unlabeled data supplying the model with more complex data inputs, which boosts the ability of the model to learn intra- and inter-class features. Extensive experiments on ISIC2018 skin lesion dataset and the Chexpert chest disease dataset demonstrate that our approach has satisfactory performance and generalization ability on the medical image classification than other state-of-the-art semi-supervised methods. [ABSTRACT FROM AUTHOR]

Appendicitis, an infection and inflammation of the appendix is a prevalent condition in children that requires immediate treatment. Rupture of the appendix may lead to several complications, such as peritonitis and sepsis. Appendicitis is medically diagnosed using urine, blood, and imaging tests. In recent times, Artificial Intelligence and machine learning have been a boon for medicine. Hence, several supervised learning techniques have been utilized in this research to diagnose appendicitis in pediatric patients. Six heterogeneous searching techniques have been used to perform hyperparameter tuning and optimize predictions. These are Bayesian Optimization, Hybrid Bat Algorithm, Hybrid Self-adaptive Bat Algorithm, Firefly Algorithm, Grid Search, and Randomized Search. Further, nine classification metrics were utilized in this study. The Hybrid Bat Algorithm technique performed the best among the above algorithms, with an accuracy of 94% for the customized APPSTACK model. Five explainable artificial intelligence techniques have been tested to interpret the results made by the classifiers. According to the explainers, length of stay, means vermiform appendix detected on ultrasonography, white blood cells, and appendix diameter were the most crucial markers in detecting appendicitis. The proposed system can be used in hospitals for an early/quick diagnosis and to validate the results obtained by other diagnostic modalities. [ABSTRACT FROM AUTHOR]

It is essential to elucidate the molecular mechanisms underlying liver transplant tolerance and rejection. In cases of mouse liver transplantation between inbred strains, immunological rejection of the allograft is reduced with spontaneous apoptosis without immunosuppressive drugs, which differs from the actual clinical result. This may be because inbred strains are genetically homogeneous and less heterogeneous than others. We exploited outbred CD1 mice, which show highly heterogeneous genotypes among individuals, to search for biomarkers related to immune responses and to construct a model for predicting the outcome of liver allografting. Of the 36 mice examined, 18 died within 3 weeks after transplantation, while the others survived for more than 6 weeks. Whole-exome sequencing of the 36 donors revealed more than 9 million variants relative to the C57BL/6 J reference. We selected 6517 single-nucleotide and indel variants and performed machine learning to determine whether or not we could predict the prognosis of each genotype. Models were built by both deep learning with a one-dimensional convolutional neural network and linear classification and evaluated by leave-one-out cross-validation. Given that one short-lived mouse died early in an accident, the models perfectly predicted the outcome of all individuals, suggesting the importance of genotype collection. In addition, linear classification models provided a list of loci potentially responsible for these responses. The present methods as well as results is likely to be applicable to liver transplantation in humans. [ABSTRACT FROM AUTHOR]

Summary Artificial Intelligence (AI) holds great potential in leveraging complex relationships within large datasets. The primary challenge lies in identifying relevant features in abstract data such as molecules, compounds or spectra. Models are trained and tested using supervised or unsupervised learning methods. Active learning and iterative processes enhance accuracy. AI offers the potential for precise predictions and innovation in both research and industry. [ABSTRACT FROM AUTHOR]