Heterogeneous Datasets Research Articles

Comparison of state-of-the-art deep learning architectures for detection of freezing of gait in Parkinson's disease.

Freezing of gait (FOG) is one of the most debilitating motor symptoms experienced by patients with Parkinson's disease (PD). FOG detection is possible using acceleration data from wearable sensors, and a convolutional neural network (CNN) is often used to determine the presence of FOG epochs. We compared the performance of a standard CNN for the detection of FOG with two more complex networks, which are well suited for time series data, the MiniRocket and the InceptionTime. We combined acceleration data of people with PD across four studies. The final data set was split into a training (80%) and hold-out test (20%) set. A fifth study was included as an unseen test set. The data were windowed (2 s) and five-fold cross-validation was applied. The CNN, MiniRocket, and InceptionTime models were evaluated using a receiver operating characteristic (ROC) curve and its area under the curve (AUC). Multiple sensor configurations were evaluated for the best model. The geometric mean was subsequently calculated to select the optimal threshold. The selected model and threshold were evaluated on the hold-out and unseen test set. A total of 70 participants (23.7 h, 9% FOG) were included in this study for training and testing, and in addition, 10 participants provided an unseen test set (2.4 h, 11% FOG). The CNN performed best (AUC = 0.86) in comparison to the InceptionTime (AUC = 0.82) and MiniRocket (AUC = 0.76) models. For the CNN, we found a similar performance for a seven-sensor configuration (lumbar, upper and lower legs and feet; AUC = 0.86), six-sensor configuration (upper and lower legs and feet; AUC = 0.87), and two-sensor configuration (lower legs; AUC = 0.86). The optimal threshold of 0.45 resulted in a sensitivity of 77% and a specificity of 58% for the hold-out set (AUC = 0.72), and a sensitivity of 85% and a specificity of 68% for the unseen test set (AUC = 0.90). We confirmed that deep learning can be used to detect FOG in a large, heterogeneous dataset. The CNN model outperformed more complex networks. This model could be employed in future personalized interventions, with the ultimate goal of using automated FOG detection to trigger real-time cues to alleviate FOG in daily life.

Semantic Modelling Approach for Safety-Related Traffic Information Using DATEX II

The significance of Linked Open Data datasets for traffic information extends beyond just including open traffic data. It incorporates links to other relevant thematic datasets available on the web. This enables federated queries across different data platforms from various countries and sectors, such as transport, geospatial, environmental, weather, and more. Businesses, researchers, national operators, administrators, and citizens at large can benefit from having dynamic traffic open data connected to heterogeneous datasets across Member States. This paper focuses on the development of a semantic model that enhances the basic service to access open traffic data through a LOD-enhanced Traffic Information System in alignment with the ITS Directive (2010/40/EU). The objective is not limited to just viewing or downloading data but also to improve the extraction of meaningful information and enable other types of services that are only achievable through LOD. By structuring the information using the RDF format meant for machines and employing SPARQL for querying, LOD allows for comprehensive and unified access to all datasets. Considering that the European standard DATEX II is widely used in many priority areas and services mentioned in the ITS Directive, LOD DATEX II was developed as a complementary approach to DATEX II XML. This facilitates the accessibility and comprehensibility of European traffic data and services. As part of this development, an ontological model called dtx_srti, based on the DATEX II Ontology, was created to support these efforts.

Heterogeneous Ship Data Classification with Spatial–Channel Attention with Bilinear Pooling Network

The classification of ship images has become a significant area of research within the remote sensing community due to its potential applications in maritime security, traffic monitoring, and environmental protection. Traditional monitoring methods like the Automated Identification System (AIS) and the Constant False Alarm Rate (CFAR) have their limitations, such as challenges with sea clutter and the problem of ships turning off their transponders. Additionally, classifying ship images in remote sensing is a complex task due to the spatial arrangement of geospatial objects, complex backgrounds, and the resolution limitations of sensor platforms. To address these challenges, this paper introduces a novel approach that leverages a unique dataset termed Heterogeneous Ship data and a new technique called the Spatial–Channel Attention with Bilinear Pooling Network (SCABPNet). First, we introduce the Heterogeneous Ship data, which combines Synthetic Aperture Radar (SAR) and optical satellite imagery, to leverage the complementary features of the SAR and optical modalities, thereby providing a richer and more-diverse set of features for ship classification. Second, we designed a custom layer, called the Spatial–Channel Attention with Bilinear Pooling (SCABP) layer. This layer sequentially applies the spatial attention, channel attention, and bilinear pooling techniques to enhance the feature representation by focusing on extracting informative and discriminative features from input feature maps, then classify them. Finally, we integrated the SCABP layer into a deep neural network to create a novel model named the SCABPNet model, which is used to classify images in the proposed Heterogeneous Ship data. Our experiments showed that the SCABPNet model demonstrated superior performance, surpassing the results of several state-of-the-art deep learning models. SCABPNet achieved an accuracy of 97.67% on the proposed Heterogeneous Ship dataset during testing. This performance underscores SCABPNet’s capability to focus on ship-specific features while suppressing background noise and feature redundancy. We invite researchers to explore and build upon our work.

Evaluation of mediastinal lymph node segmentation of heterogeneous CT data with full and weak supervision

Accurate lymph node size estimation is critical for staging cancer patients, initial therapeutic management, and assessing response to therapy. Current standard practice for quantifying lymph node size is based on a variety of criteria that use uni-directional or bi-directional measurements. Segmentation in 3D can provide more accurate evaluations of the lymph node size. Fully convolutional neural networks (FCNs) have achieved state-of-the-art results in segmentation for numerous medical imaging applications, including lymph node segmentation. Adoption of deep learning segmentation models in clinical trials often faces numerous challenges. These include lack of pixel-level ground truth annotations for training, generalizability of the models on unseen test domains due to the heterogeneity of test cases and variation of imaging parameters. In this paper, we studied and evaluated the performance of lymph node segmentation models on a dataset that was completely independent of the one used to create the models. We analyzed the generalizability of the models in the face of a heterogeneous dataset and assessed the potential effects of different disease conditions and imaging parameters. Furthermore, we systematically compared fully-supervised and weakly-supervised methods in this context. We evaluated the proposed methods using an independent dataset comprising 806 mediastinal lymph nodes from 540 unique patients. The results show that performance achieved on the independent test set is comparable to that on the training set. Furthermore, neither the underlying disease nor the heterogeneous imaging parameters impacted the performance of the models. Finally, the results indicate that our weakly-supervised method attains 90%− 91% of the performance achieved by the fully supervised training.

Aggregation of monitoring datasets for functional diversity estimation

Long-term monitoring data is central for the analysis of biodiversity change and its drivers. Time series allow a more accurate evaluation of diversity indices, trait identification and community turnover. However, evaluating data collected across different monitoring programs remains complicated because of data discrepancies and inconsistencies. Here we propose a method for aggregating datasets using diffusion maps. The method is illustrated by aggregating long-term phytoplankton abundance data from the Wadden Sea and Southern North Sea gathered by two institutions located in Germany and The Netherlands. The aggregated data allowed us to infer species traits, to reconstruct the main trait axis which drives community functionality, ultimately quantifying functional diversity of the individual samples, having used only the co-occurrence of species in samples. Although functional diversity varies greatly among sampling stations, we detect a slight positive trend in German stations, which contrasts with the clear decreasing trend observed in most of the Dutch Wadden Sea stations. At the Terschelling transect, in Southern North Sea, the stations also showed contrasting estimations of functional diversity between off-shore and in-shore stations. Our research provides further evidence that traits and functional diversity can be robustly reconstructed from monitoring data alone, showing that data aggregation can increase the accuracy of this reconstruction, being able to aggregate heterogeneous datasets.

Comparison Efficacy of VGG16 and VGG19 Insect Classification Models

This study compares two popular deep-learning models, VGG16 and VGG19, for insect classification. This study aims to evaluate insect detection architectures to automate insect identification. We use a large, heterogeneous dataset of insect species, including common pests and beneficial insects, and their images to achieve this goal. The dataset was used to re-adjust the VGG16 and VGG19 models and analyze their classification performance. With an average improvement of 1,8%, VGG19 outperforms VGG16 in insect classification accuracy. VGG19 is more robust because it can handle complex traits and subtle insect morphology differences. Each architecture's model training duration and computational resources are examined for their practicality in real-world scenarios. This study emphasizes deep learning models in insect classification and shows VGG19's higher accuracy and robustness than VGG16. These findings matter to entomologists, agricultural researchers, and pest control experts. They can improve insect identification accuracy and effectiveness using VGG19-based models, which can help solve insect-related problems in various fields.96.28 percent accuracy was achieved with the implementation of VGG16, according to the experimental findings; 97.07 percent accuracy was obtained with the implementation of VGG19.

Prediction of brain sex from EEG: using large-scale heterogeneous dataset for developing a highly accurate and interpretable ML model

This study presents a comprehensive examination of sex-related differences in resting-state electroencephalogram (EEG) data, leveraging two different types of machine learning models to predict an individual's sex. We utilized data from the Two Decades-Brainclinics Research Archive for Insights in Neurophysiology (TDBRAIN) EEG study, affirming that gender prediction can be attained with noteworthy accuracy. The best performing model achieved an accuracy of 85% and an ROC AUC of 89%, surpassing all prior benchmarks set using EEG data and rivaling the top-tier results derived from fMRI studies. A comparative analysis of LightGBM and Deep Convolutional Neural Network (DCNN) models revealed DCNN's superior performance, attributed to its ability to learn complex spatial-temporal patterns in the EEG data and handle large volumes of data effectively. Despite this, interpretability remained a challenge for the DCNN model. The LightGBM interpretability analysis revealed that the most important EEG features for accurate sex prediction were related to left fronto-central and parietal EEG connectivity. We also showed the role of both low (delta and theta) and high (beta and gamma) activity in the accurate sex prediction. These results, however, have to be approached with caution, because it was obtained from a dataset comprised largely of participants with various mental health conditions, which limits the generalizability of the results and necessitates further validation in future studies. . Overall, the study illuminates the potential of interpretable machine learning for sex prediction, alongside highlighting the importance of considering individual differences in prediction sex from brain activity.

Which Convolutional Neural Network Architecture is the Best at Detecting Alzheimer’s Disease: A head‐to‐head Comparison Using Relevance Maps

AbstractBackgroundAlzheimer’s disease (AD) is a neurodegenerative disorder which causes gradual and irreversible damage to the brain. These patterns of brain damage can be detected using T1‐weighted MRI scans by modern neuroimaging methods. Recent developments in convolutional neural networks (CNN) have achieved promising results in various computer‐vision tasks. Feature attribution methods such as layer‐wise relevance propagation (LRP) allow tracing back the information flow in CNNs to derive relevance heatmaps, which approximate the contribution of the input image regions on the model decision. We addressed the open question, which of the most common CNN architectures is best suited for medical image detection, i.e. AD classification based on MRI data.MethodFour CNN architectures are widely used in the literature: AlexNet, VGG, ResNet, and DenseNet. We adapted these CNN architectures to be used with 3D brain MRI data and trained the models on a heterogeneous dataset with N>2200 from four large studies. We applied tenfold cross‐validation and additionally evaluated the results on an independent test dataset.ResultThe more complex CNN architectures DenseNet and ResNet with ‘skip connections’, provided the best results for both group separation tasks of ‐ AD vs Cognitively Normal (CN), and Mild Cognitive Impairment (MCI) vs CN, although the overall differences in accuracy did not reach statistical significance. From the mean relevance maps obtained using the LRP relevance propagation method (see Fig. 1.), we found that DenseNet focused primarily on the medial temporal lobe and posterior cingulate cortex atrophy. ResNet’s mean relevance maps were more noisy and heterogeneous.ConclusionDenseNet provided the most focused relevance maps, best matching apriori expectations of brain regions contributing to the detection of AD. Its ‘dense connections’ between the layers enabled a highly efficient information flow at various scales. Our evaluation setup also demonstrated the value of a holistic evaluation of models, where relevance maps are used in combination with classical performance metrics.

Which Convolutional Neural Network Architecture is the Best at Detecting Alzheimer’s Disease: A head‐to‐head Comparison Using Relevance Maps

AbstractBackgroundAlzheimer’s disease (AD) is a neurodegenerative disorder which causes gradual and irreversible damage to the brain. These patterns of brain damage can be detected using T1‐weighted MRI scans by modern neuroimaging methods. Recent developments in convolutional neural networks (CNN) have achieved promising results in various computer‐vision tasks. Feature attribution methods such as layer‐wise relevance propagation (LRP) allow tracing back the information flow in CNNs to derive relevance heatmaps, which approximate the contribution of the input image regions on the model decision. We addressed the open question, which of the most common CNN architectures is best suited for medical image detection, i.e. AD classification based on MRI data.MethodFour CNN architectures are widely used in the literature: AlexNet, VGG, ResNet, and DenseNet. We adapted these CNN architectures to be used with 3D brain MRI data and trained the models on a heterogeneous dataset with N>2200 from four large studies. We applied tenfold cross‐validation and additionally evaluated the results on an independent test dataset.ResultThe more complex CNN architectures DenseNet and ResNet with ‘skip connections’, provided the best results for both group separation tasks of ‐ AD vs Cognitively Normal (CN), and Mild Cognitive Impairment (MCI) vs CN, although the overall differences in accuracy did not reach statistical significance. From the mean relevance maps obtained using the LRP relevance propagation method (see Fig. 1.), we found that DenseNet focused primarily on the medial temporal lobe and posterior cingulate cortex atrophy. ResNet’s mean relevance maps were more noisy and heterogeneous.ConclusionDenseNet provided the most focused relevance maps, best matching apriori expectations of brain regions contributing to the detection of AD. Its ‘dense connections’ between the layers enabled a highly efficient information flow at various scales. Our evaluation setup also demonstrated the value of a holistic evaluation of models, where relevance maps are used in combination with classical performance metrics.

Automated quality control of structural brain MRI scans from aging and dementia datasets

AbstractBackgroundDementias platform UK (DPUK) currently hosts over 40 dementia cohorts and provides a remote analysis environment for multimodal cross‐cohort data analysis. Nineteen cohorts include brain MRI data, hence automated quality control (QC) becomes crucial for the reliability of image analyses. This work aims to provide a framework for quality prediction of T1‐weighted (T1w) scans from aging and dementia datasets: we compared manual quality ratings with the output of two existing automated QC pipelines and propose a new QC classifier.MethodWe used 2438 T1w scans (including patients and controls, 1.5T and 3T scanners, 3 manufacturers, 11 sites) from 4 different aging and dementia datasets (ADNI, Whitehall II, Oxford Parkinson’s Disease Centre, Oxford Brain Health Clinic). Manual quality ratings for all images were performed by the dataset owners. Each image was processed in MRIQC (Esteban et al. 2017, PLoS ONE) and CAT12 (Computational Anatomy Toolbox, Gaser et al. 2022, bioRxiv). We compared the quality agreement for MRIQC predictions and CAT12’s weighted image quality ratings (IQR) with manual QC using inter‐rater reliability. Further, we explored the effect of changing the accept‐reject threshold from automated QC tools on the reliability measure. Finally, we designed a custom QC classifier (manual QC as target) by combining MRIQC and CAT12’s quality metrics as features and training linear support vector machines in a nested cross‐validation framework.ResultWe found overall agreement between automated QC methods and manual QC when applied to dementia datasets (MRIQC: Kappa = 0.30, p = 0.001; CAT12: Kappa = 0.28, p = 0.001). Adjusting the acceptance thresholds of the automated tools within dataset improved the agreement. On the combined data (training N = 1951; test N = 487), the proposed classifier showed 94.2% accuracy (sensitivity 94.3%, specificity 85.7%) on the test data. Leaving one site out (training with 10 sites, N = 2055), we found 86.4% accuracy (sensitivity 86.2%, specificity 100%) on the test site (N = 383).ConclusionThe performance of the proposed classifier appears promising on these heterogeneous datasets from different scanners. In the future, we aim to utilise this framework for improving generalisability of prediction and release the classifier on the DPUK data portal to robustly QC T1w scans for other aging and dementia cohorts.

Finetuning Pretrained Model with Embedding of Domain and Language Information for ASR of Very Low-Resource Settings

This study investigates the effective incorporation of meta-information such as domain and language in finetuning a pretrained model based on self-supervised learning (SSL) for automatic speech recognition (ASR) in very low-resource settings. SSL pretrained models have been shown to achieve comparable or even better performance to conventional end-to-end systems even when we finetune them with a small dataset. However, it still requires the specific target dataset with a considerable amount of labeled data, like 10 h, to achieve satisfactory performance. Thus, we propose to exploit heterogeneous datasets which are partially matched either in language or domain and apply multi-task learning (MTL) or adversarial learning (ADV) using the meta-information. The finetuning comprises (1) domain adaptation, which uses in-domain multi-lingual datasets, and (2) language adaptation, which uses datasets of the same language but different domains. The auxiliary task is domain identification for language adaptation and language identification for domain adaptation. We then embed the output of the auxiliary task into the encoder output of the ASR task. The target dataset is the Khmer corpus of the Extraordinary Chambers in the Courts of Cambodia (ECCC) in various sizes from one hour to 10 h. The experimental evaluations demonstrate that fusing the meta-information in MTL or ADV significantly improves ASR accuracy. Moreover, a two-step adaptation method which first conducts domain adaptation and then language adaptation is the most effective. We also show that the target labeled dataset of only 5 h gives an almost saturated performance.

A privacy-preserving federated learning with a secure collaborative for malware detection models using Internet of Things resources

The rise in malware attacks presents a growing security threat in today’s digital landscape, emphasizing the need for advanced malware detection methodology. While numerous detection systems utilizing methodologies like machine learning (ML) have been developed, they often fail to detect new and unknown malware in real time. This is because their malware datasets are generated with a very limited source and are not dynamic for up-to-date malware. Thus, with of the growth of the Internet of Things, malware have also increasingly grown, which makes current malware detection systems outdated in a short time. Therefore, IoT device resources can be exploited for building malware datasets having various types of malware families. In addition, the approach of collecting data from IoT devices and training them in a centralized cloud (CC) server is no longer acceptable due to this approach incurs significant security and privacy issues. This paper proposes PPFL-SC, an efficient privacy-preserving federated learning with a secure collaborative supporting verification. A group-oblivious signcryption cryptography is designed to be used as a privacy-preserving technique for federated learning. The CC server aggregates data without compromising individual data privacy. Moreover, the Stackelberg incentive model is designed as an incentive mechanism for encouraging IoT devices to collaborate through their heterogeneous dataset in a malware detection model. Thus, the proposed incentive mechanism is capable of increasing the participation of IoT devices and decreasing their dropping out during model updating. A security analysis confirms that PPFL-SC fulfills all security requirements for privacy-preserving federated learning. Real-world dataset experiments validate the efficiency and practicality of the proposed PPFL-SC.

Clustering pipeline for vehicle behavior in smart villages

Smart cities and villages present a plethora of opportunities for fusing and managing multi-source data. However, in the analysis of mobility patterns, the use of only one data source (i.e., road sensors) without considering other contextual data sources, limits the understanding of the process. To address this gap, we propose a pipeline that integrates multiple data sources, providing valuable information for pattern extraction, mainly based on vehicle mobility behavior and provenance. Our research also highlights the critical role of selecting the appropriate normalization algorithm to scale input features from heterogeneous data sources, which has not received sufficient attention in the literature. We conducted our analysis using data from four License Plate Recognition (LPR) cameras, spanning nine months, and incorporating several databases that include provenance, gross income, and holiday information, resulting in a dataset of over 50,000 vehicles. Using this data and our clustering pipeline, we identified various traffic patterns among residents and visitors in a rural touristic area. Our findings assist data analysts in choosing algorithms for analyzing heterogeneous datasets. Moreover, policymakers could use our results to adjust the resources, such as new parking zones.

ScDREAMER for atlas-level integration of single-cell datasets using deep generative model paired with adversarial classifier

Integration of heterogeneous single-cell sequencing datasets generated across multiple tissue locations, time, and conditions is essential for a comprehensive understanding of the cellular states and expression programs underlying complex biological systems. Here, we present scDREAMER (https://github.com/Zafar-Lab/scDREAMER), a data-integration framework that employs deep generative models and adversarial training for both unsupervised and supervised (scDREAMER-Sup) integration of multiple batches. Using six real benchmarking datasets, we demonstrate that scDREAMER can overcome critical challenges including skewed cell type distribution among batches, nested batch-effects, large number of batches and conservation of development trajectory across batches. Our experiments also show that scDREAMER and scDREAMER-Sup outperform state-of-the-art unsupervised and supervised integration methods respectively in batch-correction and conservation of biological variation. Using a 1 million cells dataset, we demonstrate that scDREAMER is scalable and can perform atlas-level cross-species (e.g., human and mouse) integration while being faster than other deep-learning-based methods.

Predicting Infectious Disease Outbreaks with Machine Learning and Epidemiological Data

Over the past several years, there has been a notable shift in the international public health arena, mostly driven by the use of machine learning methodologies and epidemiological data for the purpose of forecasting and controlling outbreaks of infectious diseases. This study explores the changing paradigm of disease outbreak prediction by examining current advancements and emerging patterns in the field of machine learning and epidemiology. In this paper, we explore the complex procedure of forecasting infectious disease outbreaks, a task of significant significance for global public health authorities. This paper examines the crucial role of machine learning algorithms in this undertaking, elucidating their capacity to analyze extensive and heterogeneous datasets in order to produce significant insights and predictions Our inquiry spans multiple facets of this complex topic. This study examines the transformative impact of machine learning models, namely deep learning and ensemble approaches, on the field. The individuals in question have exhibited remarkable proficiency in recognizing patterns, establishing correlations, and formulating predictions by utilizing past data. Consequently, this has greatly contributed to the prompt identification and readiness for potential outbreaks. Moreover, our study involves the incorporation of epidemiological data, including case reports, genetic sequencing, and population dynamics, into the machine learning architecture. This study investigates the enhanced predictive accuracy and improved comprehension of disease dynamics resulting from the integration of data-driven models and expert knowledge from the field of epidemiology. The integration of different approaches is of utmost importance when it comes to effectively tackling the distinct characteristics and problems presented by diverse infectious illnesses. Additionally, the research emphasizes the significance of incorporating a wide range of data sources, including not only data related to human health, but also environmental factors, socio-economic metrics, and patterns of human mobility. Non-conventional data sources provide essential contextual information for comprehending the dynamics of disease transmission, hence enhancing the robustness and comprehensiveness of forecasts.

PersA-FL: personalized asynchronous federated learning

We study the personalized federated learning problem under asynchronous updates. In this problem, each client seeks to obtain a personalized model that simultaneously outperforms local and global models. We consider two optimization-based frameworks for personalization: (i) Model-Agnostic Meta-Learning (MAML) and (ii) Moreau Envelope (ME). MAML involves learning a joint model adapted for each client through fine-tuning, whereas ME requires a bi-level optimization problem with implicit gradients to enforce personalization via regularized losses. We focus on improving the scalability of personalized federated learning by removing the synchronous communication assumption. Moreover, we extend the studied function class by removing boundedness assumptions on the gradient norm. Our main technical contribution is a unified proof for asynchronous federated learning with bounded staleness that we apply to MAML and ME personalization frameworks. For the smooth and non-convex functions class, we show the convergence of our method to a first-order stationary point. We illustrate the performance of our method and its tolerance to staleness through experiments for classification tasks over heterogeneous datasets.

Identifying Depression Through Machine Learning Analysis of Omics Data: Scoping Review.

Depression is one of the most common mental disorders that affects >300 million people worldwide. There is a shortage of providers trained in the provision of mental health care, and the nursing workforce is essential in filling this gap. The diagnosis of depression relies heavily on self-reported symptoms and clinical interviews, which are subject to implicit biases. The omics methods, including genomics, transcriptomics, epigenomics, and microbiomics, are novel methods for identifying the biological underpinnings of depression. Machine learning is used to analyze genomic data that includes large, heterogeneous, and multidimensional data sets. This scoping review aims to review the existing literature on machine learning methods for omics data analysis to identify individuals with depression, with the goal of providing insight into alternative objective and driven insights into the diagnostic process for depression. This scoping review was reported following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines. Searches were conducted in 3 databases to identify relevant publications. A total of 3 independent researchers performed screening, and discrepancies were resolved by consensus. Critical appraisal was performed using the Joanna Briggs Institute Critical Appraisal Checklist for Analytical Cross-Sectional Studies. The screening process identified 15 relevant papers. The omics methods included genomics, transcriptomics, epigenomics, multiomics, and microbiomics, and machine learning methods included random forest, support vector machine, k-nearest neighbor, and artificial neural network. The findings of this scoping review indicate that the omics methods had similar performance in identifying omics variants associated with depression. All machine learning methods performed well based on their performance metrics. When variants in omics data are associated with an increased risk of depression, the important next step is for clinicians, especially nurses, to assess individuals for symptoms of depression and provide a diagnosis and any necessary treatment.

Exact algorithms for a parallel machine scheduling problem with workforce and contiguity constraints

We address a real-world scheduling problem where the objective is to allocate a set of tasks to a set of machines and to a set of workers in such a way that the total weighted tardiness is minimized. Our case study encompasses four types of constraints: precedence, resource, eligibility, and contiguity. While the first three constraints are common in the scheduling literature, contiguity constraints, which can be defined as a form of precedence constraints that requires both a predecessor and its successor to be processed on the same machine with no intermediate jobs in-between (but idle time is allowed), have never been studied in the literature. We present four exact methods to solve the problem: two methods use integer linear programming, one uses constraint programming, and one uses a combinatorial Benders’ decomposition. We introduce method-specific strategies to model the contiguity constraints for each of the proposed methods. We empirically evaluate, through an extensive set of computational experiments, the performance of the four methods on a heterogeneous dataset composed of real, realistic, and random instances, and outline that every method offers a competitive advantage on a targeted subset of instances. We also show that our algorithms can be generalized to solve related scheduling problems with contiguity constraints.

PatchAugNet: Patch feature augmentation-based heterogeneous point cloud place recognition in large-scale street scenes

Point Cloud Place Recognition (PCPR) in street scenes is an essential task in the fields of autonomous driving, robot navigation, and urban map updating. However, the domain gap between heterogeneous point clouds and the difficulty of feature characterization in large-scale complex street scenes pose significant challenges for existing PCPR methods. Most PCPR methods only take into account point clouds collected by the same platforms and sensors, thus they are with poor domain transferability. In this paper, we propose PatchAugNet, which utilizes patch feature augmentation and adaptive pyramid feature aggregation to achieve better performance and generalizability for Heterogeneous Point Cloud-based Place Recognition (HPCPR) tasks. Firstly, multi-scale local features are extracted by the pyramid feature extraction module. Secondly, local features are enhanced by the patch feature augmentation module to overcome the domain gap problem and achieve better feature representation as well as network generalizability. Finally, a global feature is generated using an adaptive pyramid feature aggregation module, which automatically adjusts and balances the proportion of intra-scale and inter-scale features according to the scene content. To evaluate the performance of PatchAugNet, a large-scale heterogeneous point cloud dataset consisting of high-precision Mobile Laser Scanning (MLS) point clouds and helmet-mounted Portable Laser Scanning (PLS) point clouds is collected. The dataset covers various street scenes with a length of over 20 km. The comprehensive experimental results indicate that PatchAugNet achieves State-Of-The-Art (SOTA) performance with 83.43 % recall@top1% and 60.34 % recall@top1 on unseen large-scale street scenes, outperforming existing SOTA PCPR methods by + 9.57 recall@top1% and + 15.50 recall@top1, while exhibiting better generalizability. For source code and detailed experimental results, please refer to: https://github.com/WHU-USI3DV/PatchAugNet.

Enhancing Predictive Expert Method for Link Prediction in Heterogeneous Information Social Networks

Link prediction is a critical prerequisite and foundation task for social network security that involves predicting the potential relationship between nodes within a network or graph. Although the existing methods show promising performance, they often ignore the unique attributes of each link type and the impact of diverse node differences on network topology when dealing with heterogeneous information networks (HINs), resulting in inaccurate predictions of unobserved links. To overcome this hurdle, we propose the Enhancing Predictive Expert Method (EPEM), a comprehensive framework that includes an individual feature projector, a predictive expert constructor, and a trustworthiness investor. The individual feature projector extracts the distinct characteristics associated with each link type, eliminating shared attributes that are common across all links. The predictive expert constructor then creates enhancing predictive experts, which improve predictive precision by incorporating the individual feature representations unique to each node category. Finally, the trustworthiness investor evaluates the reliability of each enhancing predictive expert and adjusts their contributions to the prediction outcomes accordingly. Our empirical evaluations on three diverse heterogeneous social network datasets demonstrate the effectiveness of EPEM in forecasting unobserved links, outperforming the state-of-the-art methods.