Machine Learning Methods Research Articles

INTEGRATION OF MACHINE LEARNING METHODS INTO SEMANTIC ANALYSIS OF OSINT DATA: ALGORITHMS AND RESULTS

The aim of this study is to develop a universal software module for semantic analysis of open-source intelligence data using machine learning methods, enabling the automation of data collection, processing, and clustering of large datasets. Special emphasis is placed on improving analysis accuracy and simplifying result interpretation for users from various fields, such as cybersecurity, business analytics, and journalism. The study employs modern natural language processing methods, including Term Frequency-Inverse Document Frequency, Bidirectional Encoder Representations from Transformers for semantic text analysis, and the Density-Based Spatial Clustering of Applications with Noise algorithm for data grouping. Data collection was carried out using social media APIs, news portals, and automated web scraping tools. The module’s performance was evaluated using precision, recall, and F1-score metrics. For the first time, this study proposes the integration of deep learning models, namely Bidirectional Encoder Representations from Transformers, with clustering algorithms to address open-source intelligence tasks. The system provides flexibility in adapting to different data sources and multilingual processing. A significant achievement is the improvement of semantic analysis by considering query context, which ensures higher result relevance. The developed module can be applied in cybersecurity for threat detection, in business for competitor analysis and market monitoring, and in journalism for fact-checking, research, fake news detection, and analysis of information campaigns. The integration of multiple data sources enables a comprehensive approach to information analysis. The tested module demonstrated high efficiency, achieving a clustering accuracy of over 90 % and the capability to process large datasets (up to 10,000 documents in 3 minutes). The system successfully identifies key entities, automatically clusters data, and provides visualization in a user-friendly format.

A Hybrid Machine Learning Model for Peripheral Artery Disease Prediction and Real-Time Applications

Peripheral Artery Disease (PAD) is a common and serious circulatory problem for which early and definite diagnosis is necessary to prevent further health complications. The existing diagnostic techniques of Ankle-Brachial Index (ABI) and Doppler ultrasound have several disadvantages: the examination is operator-dependent, the processing times are long, and may be inapplicable or have adaptability in complicated cases. The use of Machine Learning (ML) techniques, such as Support Vector Machines and Random Forest, to overcome these issues, may face problems handling real-time application and non-homogeneous data. The research at hand overcomes such challenges by proposing a Hybrid ML Algorithm for Peripheral Artery Prediction (HMAPAP) based on GBM combined with LSTM networks. The proposed method improved the diagnostic accuracy by 0.25%, processing efficiency by 0.20%, and real-time adaptability by 0.30%, of the traditional ML methods.

A review on learning-based algorithms for tractography and human brain white matter tracts recognition.

Human brain fiber tractography using diffusion magnetic resonance imaging is a crucial stage in mapping brain white matter structures, pre-surgical planning, and extracting connectivity patterns. Accurate and reliable tractography, by providing detailed geometric information about the position of neural pathways, minimizes the risk of damage during neurosurgical procedures. Both tractography itself and its post-processing steps such as bundle segmentation are usually used in these contexts. Many approaches have been put forward in the past decades and recently, multiple data-driven tractography algorithms and automatic segmentation pipelines have been proposed to address the limitations of traditional methods. Several of these recent methods are based on learning algorithms that have demonstrated promising results. In this study, in addition to introducing diffusion MRI datasets, we review learning-based algorithms such as conventional machine learning, deep learning, reinforcement learning and dictionary learning methods that have been used for white matter tract, nerve and pathway recognition as well as whole brain streamlines or whole brain tractogram creation. The contribution is to discuss both tractography and tract recognition methods, in addition to extending previous related reviews with most recent methods, covering architectures as well as network details, assess the efficiency of learning-based methods through a comprehensive comparison in this field, and finally demonstrate the important role of learning-based methods in tractography.

Trajectory-Ordered Objectives for Self-Supervised Representation Learning of Temporal Healthcare Data Using Transformers: Model Development and Evaluation Study.

The growing availability of electronic health records (EHRs) presents an opportunity to enhance patient care by uncovering hidden health risks and improving informed decisions through advanced deep learning methods. However, modeling EHR sequential data, that is, patient trajectories, is challenging due to the evolving relationships between diagnoses and treatments over time. Significant progress has been achieved using transformers and self-supervised learning. While BERT-inspired models using masked language modeling (MLM) capture EHR context, they often struggle with the complex temporal dynamics of disease progression and interventions. This study aims to improve the modeling of EHR sequences by addressing the limitations of traditional transformer-based approaches in capturing complex temporal dependencies. We introduce Trajectory Order Objective BERT (Bidirectional Encoder Representations from Transformers; TOO-BERT), a transformer-based model that advances the MLM pretraining approach by integrating a novel TOO to better learn the complex sequential dependencies between medical events. TOO-Bert enhanced the learned context by MLM by pretraining the model to distinguish ordered sequences of medical codes from permuted ones in a patient trajectory. The TOO is enhanced by a conditional selection process that focus on medical codes or visits that frequently occur together, to further improve contextual understanding and strengthen temporal awareness. We evaluate TOO-BERT on 2 extensive EHR datasets, MIMIC-IV hospitalization records and the Malmo Diet and Cancer Cohort (MDC)-comprising approximately 10 and 8 million medical codes, respectively. TOO-BERT is compared against conventional machine learning methods, a transformer trained from scratch, and a transformer pretrained on MLM in predicting heart failure (HF), Alzheimer disease (AD), and prolonged length of stay (PLS). TOO-BERT outperformed conventional machine learning methods and transformer-based approaches in HF, AD, and PLS prediction across both datasets. In the MDC dataset, TOO-BERT improved HF and AD prediction, increasing area under the receiver operating characteristic curve (AUC) scores from 67.7 and 69.5 with the MLM-pretrained Transformer to 73.9 and 71.9, respectively. In the MIMIC-IV dataset, TOO-BERT enhanced HF and PLS prediction, raising AUC scores from 86.2 and 60.2 with the MLM-pretrained Transformer to 89.8 and 60.4, respectively. Notably, TOO-BERT demonstrated strong performance in HF prediction even with limited fine-tuning data, achieving AUC scores of 0.877 and 0.823, compared to 0.839 and 0.799 for the MLM-pretrained Transformer, when fine-tuned on only 50% (442/884) and 20% (176/884) of the training data, respectively. These findings demonstrate the effectiveness of integrating temporal ordering objectives into MLM-pretrained models, enabling deeper insights into the complex temporal relationships inherent in EHR data. Attention analysis further highlights TOO-BERT's capability to capture and represent sophisticated structural patterns within patient trajectories, offering a more nuanced understanding of disease progression.

How sudden public events shape firm digital transformation: evidence from Chinese listed firms

PurposeThe outbreak of COVID-19 made a huge impact on firms and also had an important influence on their digital development strategies. Using difference-in-differences (DID) model, this study investigates the influence of COVID-19 shock on digital transformation (DT).Design/methodology/approachThis study conducts a natural experiment based on COVID-19 shock and use machine learning methods to measure the level of digitalization of firms, and used DID model to explore the changes in the DT of firms before and after COVID-19 outbreak.FindingsThe empirical results show that the DT process of firms in the experimental group was significantly accelerated after the outbreak of COVID-19, suggesting that COVID-19 shock has catalyzed an acceleration in the DT processes within firms. In addition to assessing the direct impact of COVID-19 on DT, this study also explores the influence mechanisms through which COVID-19 shock influences firm DT. The influence mechanism test shows that COVID-19 shock has compelled firms to accelerate their DT by increasing operational costs and reducing the stability of customer relationships. On this basis, this study also systematically explored the role of optimal allocation of external resources in the process of COVID-19 shock affecting DT of firms. This study found that COVID-19 shock was more significant in enhancing firm DT in situations characterized by advanced digital finance development and advanced labor aggregation. Furthermore, heterogeneity analysis shows that COVID-19 shock promotes DT more significantly with a higher degree of industry competition, a higher share of secondary industries in the region and in more market-oriented regions.Originality/valueThis research provides new insights into dynamic process of DT catalyzed by the COVID-19, while also highlighting the critical role of external resources and market conditions in facilitating this transformation.

Advancing low-cost air quality monitor calibration with machine learning methods.

Advancing low-cost air quality monitor calibration with machine learning methods.

Using machine learning methods to predict the outcome of psychological therapies for post-traumatic stress disorder: A systematic review.

Using machine learning methods to predict the outcome of psychological therapies for post-traumatic stress disorder: A systematic review.

Optimized design of RC moment frames with machine learning methods

Optimized design of RC moment frames with machine learning methods

Protocol for assessing distances in pathway space for classifier feature sets from machine learning methods.

Protocol for assessing distances in pathway space for classifier feature sets from machine learning methods.

A systematic literature review on the use of artificial intelligence for cybercrime rate forecasting

Cybercrime has a significant impact on the quality of life and economy of individuals, businesses and countries, and the speed of the increase has made it a pressing issue in today's digital age. This systematic review aims to identify the artificial intelligence models recently developed to forecast the rate of cybercrime and to help authorities and police forces define strategies in the fight against cybercrime. The PRISMA methodology was used with 229 articles retrieved from Scopus, IEEE and Web of Science, of which 30 met the eligibility criteria. The results showed that the traditional machine learning methods random forest, support vector machine (SVM) and logistic regression (LR) excel in their use to forecast cybercrimes by achieving more accurate results among the different methods tested. It was concluded that machine learning methods are, so far, effective in forecasting the rate of cybercrime, with accuracy ratios of up to 99.9%. However, the potential for future research lies in creating new forecasting models such as autoregressive integrated moving average long short term memory (ARIMA-LSTM) proposed in this study to improve the performance and accuracy of cybercrime forecasting.

The impact of training image quality with a novel protocol on artificial intelligence-based LGE-MRI image segmentation for potential atrial fibrillation management.

The impact of training image quality with a novel protocol on artificial intelligence-based LGE-MRI image segmentation for potential atrial fibrillation management.

Contrast-Induced Acute Kidney Injury in Lower Limb Percutaneous Transluminal Angioplasty: A Machine Learning Approach for Preoperative Risk Prediction.

Contrast-Induced Acute Kidney Injury in Lower Limb Percutaneous Transluminal Angioplasty: A Machine Learning Approach for Preoperative Risk Prediction.

Research on the impact of urban built environments on PM2.5 pollution based on machine learning methods

Research on the impact of urban built environments on PM2.5 pollution based on machine learning methods

MicroRNA-mediated resistance mechanisms in NAT-treated HER2-positive breast cancer (BC): A single centre retrospective analysis from matched primary and residual tumours.

e12613 Background: HER2+ accounts for 20–25 % of all BC, is aggressive and clinically challenging. Though treatment outcomes have improved with anti-HER2 agents in the past decade, primary and acquired drug resistance limit the benefits of treatment. MicroRNAs(MIRs) regulate gene expression post-transcriptionally and influence drug resistance. This study examined miRNA-mediated resistance to anti-HER2 therapy through analysis of 78 matched primary and residual tumor specimens. Methods: 78 primary FFPE tumour specimens from HER2 + BC, treated uniformly with standard NAT were accepted for analysis after Institutional Ethics approval. Using public datasets, we identified 8 prognostically significant miRs and in our series, analysed them via Taqman RT-qPCR. A predictive regression model incorporating miR-98, 15b, 193A, and 187 was developed using machine learning methods. Results: Patients showed 54% overall pCR rate following NAT, with 65% of pCR under dual anti-HER2 blockade. HR-ve /HER2+ cases demonstrated superior response (61% vs 39% p<0.05). High miRNA expression correlated with poor response (p<0.05). Dual anti-HER2 therapy and early clinical stage significantly predicted pCR (p=0.014, p=0.037). Analysis of subset of residual tumors and comparing with primary, revealed differential increased levels of oncogenic miRs-15b (p=0.021), 193A (p=0.0003), 98 (p=0.03), 129(p=0.0001), and 130b (p=0.7), suggesting increased acquired resistance. This pattern was more evident in ER-ve HER2+ cases (p<0.05). Conclusions: We identified candidate miRNAs which predict response to NAT in HER2+ BC, and this represents understanding the epigenetic mechanisms of treatment resistance in HER2+ BC, potentially leading to more effective, personalized therapeutic strategies. More analysis of primary and matched residual tumours are ongoing with functional validation for selected MIRs underway.

Multiple bolt looseness detection using SH-typed guided waves: Integrating physical mechanism with monitoring data.

Multiple bolt looseness detection using SH-typed guided waves: Integrating physical mechanism with monitoring data.

Development and validation of a predictive machine learning model for postoperative long-term diabetes insipidus following transsphenoidal surgery for sellar lesions.

Development and validation of a predictive machine learning model for postoperative long-term diabetes insipidus following transsphenoidal surgery for sellar lesions.

Predicting survivorship and caregiver burden in multiple myeloma (MM) using machine learning (ML): Insights from a large, national, prospective study.

e13661 Background: MM survival has improved significantly, resulting in increased survivorship and caregiver burden and is crucial to understand the real-world impact of MM. We analyzed results from a large, prospective MM patient survey using ML models to identify key predictive factors for survivorship and caregiver burden. Methods: A 45-item questionnaire spanning sociodemographics, treatment preferences, symptom burden and caregiver impact was distributed to MM patients at Mayo Clinic and International Myeloma Foundation support groups. Questions about quality of life (QOL) and treatment satisfaction (TS) represented survivorship outcomes while caregiver workday loss (WL) and job loss (JL) were surrogates for caregiver burden. 6 ML models (multinomial, random forest (RF), XGB (XGBoost), k-NN, GBM, Naïve Bayes) with a training: testing cohort of 80:20 were used to predict each of the 4 outcomes independently, considering model accuracy, C-index, precision, recall, and F1 score (Table), and using R studio (V4.4.1) and Bluesky Statistics (V10.3.4). Results: 2,239 participants (54% male, 89% Whites) with median age 68 yrs completed the survey. 71% respondents had a college degree, 44% resided in Midwest states, 62% reported household income >$75,000, and 67% had Medicare/Medicaid. Median age at MM diagnosis was 67 yrs with 58% diagnosed within the past 5 yrs and 73% currently receiving MM treatment. 72% reported their spouse as the primary caregiver. XGB was the best predictive model for QOL with gender, household income, stem cell transplant, CAR-T therapy, support from medical staff, and symptom burden (pain, neuropathy, mental well-being) as significant predictors. The RF model was the best for TS with pain and neuropathy levels as the most impactful predictors. For caregiver burden, XGB was the best model for WL with gender, household income, CAR-T therapy, financial strain, lack of support from medical staff, self-workdays missed, self-job loss, and pain and neuropathy levels as significant predictors. For JL, RF was the best model with significant variables being workdays missed by self, increased healthcare needs, symptom burden (pain, neuropathy, and mental well-being), and restricted physical activity. Conclusions: Our study shows the significant impact of MM survivorship and efficiently identifies predictive factors associated with survivorship and caregiver burden from a large prospective dataset using validated ML models. These findings highlight the complex challenges from MM and provide contemporary ML methods to better understand their interplay such that a framework of real-world interventions can be developed. Outcome Best ML Model Accuracy C-Index Precision Recall F1 score QOL XG 0.996 0.996 0.996 0.997 0.996 TS RF 0.984 0.900 0.937 0.948 0.937 WL XGB 0.984 1.000 0.944 0.947 0.945 JL RF 0.989 0.800 0.829 0.848 0.835

Machine learning: An effective tool for monitoring and ensuring food safety, quality, and nutrition.

Machine learning: An effective tool for monitoring and ensuring food safety, quality, and nutrition.

Predicting the solubility of drugs in supercritical carbon dioxide using machine learning and atomic contribution.

Predicting the solubility of drugs in supercritical carbon dioxide using machine learning and atomic contribution.

Neuroimaging and machine learning in eating disorders: a systematic review

Abstract Purpose Eating disorders (EDs), including anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorder (BED), are complex psychiatric conditions with high morbidity and mortality. Neuroimaging and machine learning (ML) represent promising approaches to improve diagnosis, understand pathophysiological mechanisms, and predict treatment response. This systematic review aimed to evaluate the application of ML techniques to neuroimaging data in EDs. Methods Following PRISMA guidelines (PROSPERO registration: CRD42024628157), we systematically searched PubMed and APA PsycINFO for studies published between 2014 and 2024. Inclusion criteria encompassed human studies using neuroimaging and ML methods applied to AN, BN, or BED. Data extraction focused on study design, imaging modalities, ML techniques, and performance metrics. Quality was assessed using the GRADE framework and the ROBINS-I tool. Results Out of 185 records screened, 5 studies met the inclusion criteria. Most applied support vector machines (SVMs) or other supervised ML models to structural MRI or diffusion tensor imaging data. Cortical thickness alterations in AN and diffusion-based metrics effectively distinguished ED subtypes. However, all studies were observational, heterogeneous, and at moderate to serious risk of bias. Sample sizes were small, and external validation was lacking. Conclusion ML applied to neuroimaging shows potential for improving ED characterization and outcome prediction. Nevertheless, methodological limitations restrict generalizability. Future research should focus on larger, multicenter, and multimodal studies to enhance clinical applicability. Level of Evidence: Level IV, multiple observational studies with methodological heterogeneity and moderate to serious risk of bias.