Effect of accelerating voltages on indexing of SEM-EBSD via pattern matching.

Stringology-based motif discovery for electrophysiological time series: A framework for temporal pattern analysis with an ADHD case study.

Reassembling fragmented 3D objects is challenging, particularly for archaeological artifacts affected by surface degradation and material heterogeneity. We propose a graph-based method integrating geometric and topological features to match fragments by analyzing surface patch arrangements instead of individual shapes, improving robustness to erosion and noise. Fragments are first segmented using Reeb graph-based partitioning. From this, an adjacency graph encoding the spatial organization of surface regions is constructed. Local geometry is described using the shape index to detect complementary areas across fragments. Pairwise matches are obtained by comparing structural patterns within adjacency graphs and filtered using geometric consistency constraints. Global reassembly is achieved through confidence-guided clustering with collision checks. Experiments on two public datasets achieve 80% and 95% pairwise matching recall and enable correct reconstruction of objects with up to 62 fragments, outperforming classical and recent AI-based methods.

Professional men's tennis has transformed significantly since the Open Era began in 1968, with shifts in playing styles, competition levels, and match patterns driven by advances in technology, training, and rules. However, systematic quantification of these changes remains unexplored. This study examines approximately 198,000 ATP matches from 1968 to 2025 through network analysis and machine learning techniques, applying concept drift methods to characterize temporal changes and understand how the sport has evolved across eras and court surfaces. Key findings include greater overall competitiveness, with more frequent upsets and fewer one-sided wins, as top-player dominance has spread more evenly across the field. Matches have grown longer by about 16.5 minutes on average since 1991, serves have become more effective (with 1.7 more aces per match and 2.4% higher win rates on serve points), and return effectiveness has declined by 2.4%. Court surfaces show increasing similarity, with grass courts converging toward the pace of other surfaces. These findings quantify the sport's shift toward a more powerful, endurance-based game, amid advancements in racket technology and player conditioning. This research demonstrates the power of machine learning to uncover the evolutionary patterns of the tennis game.

Three-dimensional green quantity (3DGQ) mapping and integrated 2D-3D assessment reveal hidden urban green deficits and inequities.

Digitization has made our lives easy and fast in many aspects, but it creates many challenges for researchers and software hardware developers. Every organization, institution, industry, or operational system mandatorily maintains a consistent database with efficient data access. Managing extensive databases with increasing sizes, changing properties, and handling computationally intensive complex queries have been challenging for researchers for decades. In real-life scenarios, the size and content of the database are frequently alterable. They must be handled effectively and efficiently to maintain consistency and integrity constraints and manage the available storage capacity. Different users have several types of data retrieval requests or queries. Our investigation shows that these queries share standard bottom-most level sub-queries (join, filtering etc.). In this paper, we have proposed a novel advanced method to speed up the execution of complex, long-running queries with similar basic-level sub-queries, utilizing the highly parallel nature of GPGPU (General Purpose Graphics Processing Unit) Shuai Che et al. (2008). We have also introduced an efficient memory management scheme with parallel read-write operations using the hashing technique for faster database access. The significant advantage of our proposed mechanism is that the dependency graph from the requested query with a higher number of nodes in each level leads to higher utilization of concurrently running threads.Another advantage is that caching the lowest-level sub-query results keeps track of their consistency with changing database records, which reduces the execution time for successive queries with similar sub-queries. We have gained significant speedup and reduced computational effort by performing every task and every micro-operation in parallel or in a pipelined fashion. Using the “access frequency” tracking method in cached results and the manual “asynchronous pre-fetching” technique in Unified memory, we have significantly reduced the page-fault rates and the I/O latency bottleneck problem in GPU programming. We did a comparison study of some related previous recent works. We have plotted graphs comparing the elapsed times of three different complex nested long-running queries (which include all kind of query commands like selection, projection, join, aggregate functions, grouping, group filtering, pattern matching, sorting etc.)on five different-sized databases for CPU, GPU, and GPGPU (CPU-GPU combined system) running our proposed method with approximate 60 %, 75 %, and 90 % cache-hits on synthetic database with star schema benchmark. Our proposed mechanism outperforms other existing systems, and this gain increases with the increase in database size and the complexity of the requested queries.

ABSTRACT This study investigates complexity matching pattern between daily milk yield in dairy cows and variations in Schumann resonance frequency bands. The analysis uses 895 days of data from a dairy herd in Lithuania and examines five low-frequency bands that overlap with canonical mammalian brainwave ranges: delta (0–3.5) Hz, theta (3.5–7) Hz, alpha (7–15) Hz, beta (15–32) Hz, and gamma (32–65) Hz. A matrix-based approach using Hankel matrix and Perfect Matrix of Lagrange Differences was applied in order to identify the temporal complexity matching between the milk yield and Schumann resonance magnetic field signals. Statistical validation was also performed using Spearman rank correlations. The results indicate the complexity matching across all frequency bands, with the strongest associations observed for the alpha band. Although the correlations are modest, the findings suggest that variations in the Earth’s natural electromagnetic environment may coincide with fluctuations in dairy cow productivity.

Pseudophosphatases are proteins with mutations in theircatalyticmotifs, resulting in the predicted loss of enzymatic activity. Althoughpseudophosphatases are established regulators of various signalingpathways in humans and other metazoans, their biological roles infungi remain largely unexplored. Identifying fungal-specific pseudophosphatasesis particularly important because they control fungal growth, development,and virulence through noncatalytic mechanisms and could representselective antifungal targets due to the absence of close human homologues.Here, we present a comprehensive overview of fungal pseudophosphatasesidentified across all major phosphatase families. To compile a robustdata set, we integrated information from literature and databasesas well as systematically scanned the phosphatomes of the human pathogenicyeast Cryptococcus neoformans, theplant pathogen Fusarium graminearum, and the neglected pathogenic fungus Talaromycesmarneffei. We identified candidate pseudophosphatasesand assessed their conservation through BLAST analysis across humans,true yeasts, filamentous ascomycetes, and basidiomycetes. We classifiedpseudophosphatase candidates into 10 distinct groups, including threegroups (CC1-type Oca, CC1-type Yvh1, and HAD-type NIF) that appearto be fungal-specific with no human homologues and exhibit lineage-specificfeatures. Available evidence indicates that fungal pseudophosphatasescontribute to signaling pathways that regulate development, metabolism,stress responses, and virulence. In addition, we developed the PseudophosphataseScanner tool as a post-HMM (Hidden Markov Model) analysis pipelineto enable genome-wide pseudophosphatase detection. By combining HMMscan-based fold assignment with motif-level pattern matching, thePseudophosphatase Scanner distinguishes canonical motifs, relaxedvariants, and degenerated fold remnants, facilitating functional interpretationof phosphatase active sites at the sequence level. The PseudophosphataseScanner is accessible both as a Python script for local executionand as a Google Colab notebook, offering a point-and-click interfacefor cloud-based analysis. This study provides a catalog of fungalpseudophosphatases and a bioinformatics platform for efficient large-scalepseudophosphatase discovery.

Reversed sound symbolism in speech sound disorders: Evidence for atypical cross-modal integration in a transparent language (Turkish).

Polishing, the process of correcting base-level errors in genome assemblies, is a critical step for ensuring accuracy in downstream analyses, such as variant calling, gene annotation, and clinical genomics applications. While recent advances in long-read sequencing technologies have helped improve assembly contiguity and genome completeness, maintaining high base-level accuracy in those genomes remains challenging due to the still appreciable errors associated with certain long-read sequencing technologies. Existing polishing approaches face notable trade-offs: alignment-based methods achieve high accuracy but incur long run times, alignment-free k-mer-based tools are scalable but struggle in regions with dense errors, and machine learning-based polishers often only perform well on specific platforms and require read-to-assembly alignments. We present AIEdit, a machine learning-based polisher designed to operate alignment-free, generalizing across sequencing platforms while remaining computationally efficient. We developed AIEdit by combining spaced seed matching with a neural network trained to detect and correct dense error patterns in an alignment-free manner. We benchmarked the method on simulated and experimental DNA sequencing data. On simulated human long-read assemblies with high error rates, AIEdit reduced error rates by 58% compared to ntEdit’s 21%, completing in 2.7 hours using 230 GB of memory, faster than POLCA and Medaka (multi-day run times) and using 3 × less memory than JASPER (689 GB). On experimental Oxford Nanopore Technologies (ONT) data from the NA24385 human genome, AIEdit increased the Merqury quality score (QV) from 28.7 to 32.9 in 9.5 hours, achieving comparable accuracy to Medaka (QV 32.7) in a fraction of the time (1.5 + days) and outperforming k-mer-based tools ntEdit (QV 31.0) and JASPER (QV 31.7). Overall, AIEdit enables scalable and accurate genome polishing across diverse datasets.

Reliably identifying relevant brain areas implicated by the simulated activity from cognitive models is still an unsolved problem for cognitive modeling, particularly when matching model output with human electroencephalography (EEG) data. We propose a new method involving postprocessing of ACT-R module activity and clustered EEG component activity, and performing generalized least squares analysis to find matching patterns between predicted and observed data, thereby inferring neural substrates of distinct cognitive processes. This approach holds several advantages over other methods by controlling for autocorrelation and unequal variances. To exemplify its application, we used a cognitive model and EEG data from a mental spatial transformation study to show how this method finds areas involved in representational and transformational spatial processing. Parietal areas involved with spatial activity were identified, in line with prior studies on spatial cognition. In addition, previously established associations between ACT-R and brain areas were confirmed. Finally, we discuss limitations and possibilities of the approach.

This study aims to analyze the role of strategic management in improving the quality of Qur'an memorization in tahfidz programs at Islamic junior high schools (Madrasah Tsanawiyah) in Purbalingga from the perspective of the sociology of educational institutions. The research employs a qualitative approach with a multiple case study design at MTs Muhammadiyah 01 Purbalingga and MTs Ma’arif NU 05 Majasari. Data were collected through interviews, observations, documentation, and archival records, and analyzed using pattern matching, explanation building, and cross-case synthesis techniques. The findings reveal that the implementation of strategic management—covering strategy formulation, implementation, and evaluation—plays a significant role in enhancing the effectiveness of tahfidz programs. Strategy formulation through SWOT and PESTLE analysis helps schools comprehensively understand internal and external conditions. Strategy implementation supported by resource management and supervision using the Balanced Scorecard improves program execution. Meanwhile, periodic evaluation through monitoring systems and feedback loops enables continuous improvement in students’ memorization achievements. Overall, comprehensive strategic management enhances the quality of Qur'an memorization and fosters competitive advantage in Islamic education institutions.

Domain Generation Algorithms (DGAs) pose a persistent threat by enabling malware to dynamically generate numerous command-and-control domains, evading traditional blocklists. While machine learning-based detectors have achieved high accuracy, they operate as statistical pattern matchers and lack the human-like anomaly perception that enables security experts to intuitively recognize unnatural domains. This paper introduces CogNormDGA, a cognitive-driven framework that models normal domain characteristics from a defender’s perspective while also anticipating how attackers might exploit cognitive blind spots. Inspired by dual-process theory, CogNormDGA combines intuitive, pattern-based screening (System 1) with analytical, rule-based evaluation of phonotactic, morphological, and semantic violations (System 2). The cognitive principles of System 1 and System 2 are computationally realized as two distinct pathways: an Attentional Salience Network and a Linguistic Constraint Evaluator, respectively. The framework produces interpretable outputs via attention saliency maps and cognitive violation reports. Extensive experiments on 400,000 domains spanning 33 DGA families demonstrate that CogNormDGA achieves competitive detection performance (F1-score 0.941) while establishing a cognitive-driven detection paradigm that produces human-aligned explanations—a property critical for practical security. It shows promising results on low-entropy and novel DGA families. Human subject studies confirm strong alignment between the model’s internal explanations and expert reasoning. Furthermore, CogNormDGA is particularly effective against low-entropy DGA families that exploit cognitive blind spots. By bridging cognitive science and cybersecurity, our work offers an interpretable and human-aligned approach to threat detection, with promising resilience that requires further validation.

Elastic Degenerate (ED) strings and Elastic Founder (EF) graphs, here collectively named variable strings, are two representations of acyclic components of pangenomes which extend the well-known notion of indeterminate string. Recent studies have focused extensively on algorithmic tasks involving these structures and other forms of variable strings that they generalize. Among such tasks, the basic operation of matching a pattern into a text, a fundamental toolkit for pangenomic data analysis, deserves special attention. In this paper, (1) we establish a clear taxonomy across ED strings and EF graphs, categorizing types of variable strings from the simplest linear (solid) string to the most complex general cases; (2) we consider the problem match(X,Y) of matching a solid or variable pattern of type X into a variable text of type Y, and investigate its time complexity when X and Y are chosen from types of variable strings in the taxonomy of (1). For all possible X and Y, we either provide a non-trivial, often sub-quadratic, upper bound for match(X,Y), or we prove a quadratic conditional lower bound, taking as a reference the existing quadratic conditional lower bounds for match(solid,ED) and match(solid,EF). A preliminary version of this work appeared in [Ascone et al., WABI 2024].

Abstract SQL Injection (SQLi) continues to rank among the most exploited vulnerabilities in web applications globally, appearing in the OWASP Top 10 for over two decades. This paper presents SQLiShield, a novel multi-layer detection and prevention middleware engineered for Python/Flask web applications. SQLiShield fuses three hierarchical protection layers: a deterministic engine of 53 pre-compiled regular expression patterns (Layer 1), an adaptive Random Forest ML classifier trained on 6,941 labeled payload samples (Layer 2), and a structural input sanitization fallback (Layer 3). The framework is deployed as a Flask before_request hook, transparently intercepting all GET parameters, POST form fields, JSON bodies, and HTTP cookies before they reach application logic. Empirical evaluation using the Kaggle SQL Injection Dataset demonstrates that the combined multi-layer architecture achieves 99.2% detection accuracy, 99.5% precision, 98.8% recall, and a 0.5% false positive rate — substantially outperforming standalone regex (87.0%, 18% FPR) and standalone ML (97.3%, 1.8% FPR). End-to-end detection latency averages 3.7 ms per request. Automated penetration testing with SQLMap across 100 obfuscated payloads yielded zero successful bypasses. A companion Flask demonstration application exposes four intentionally vulnerable endpoints — authentication bypass, UNION-based extraction, secondorder injection, and a real-time admin dashboard — enabling controlled side-by-side comparison of protected and unprotected execution paths. Keywords: SQL Injection, SQLiShield, Random Forest, Web Application Firewall, Flask Middleware, OWASP, Intrusion Detection, Regex Pattern Matching, Parameterized Queries, Defense-in-Depth, Parul University

In this research, we conducted a systematic review of artificial intelligence techniques used for the diagnosis of lung cancer. A systematic search of Web of Science, PubMed, Scopus, Epistemonikos, Cochrane, Medline, and Embase databases was carried out, containing the literature published up to June 2025. Prediction model risk of bias assessment tool (PROBAST) was used to evaluate the risk of bias and applicability of the diagnostic model studies included in the current research. 204 studies were included. The included articles utilized various AI techniques, including CNN (Convolutional Neural Network), SVM (Support Vector Machine), RF (Random Forest), KNN (K-Nearest Neighbor), PM-DL (Pattern Matching combined with Deep Learning), ANN (Artificial Neural Network), DNN (Deep Neural Network), CDNs (Convolutional Dense Networks), DLS (Deep Learning System), LSTM (Long Short-Term Memory), NNE (Neural Network Ensemble), and LDA (Linear Discriminant Analysis). The CNN model appears to be the most commonly used model in the papers. It was observed that applying deep learning models to preprocessed and augmented medical images led to improved performance metrics, including AUC, sensitivity, and accuracy. The accuracy of artificial intelligence techniques ranged from 68.4 to 100, while the sensitivity varied from 50.0 to 100. The specificity of the artificial intelligence techniques ranged from 50.0 to 100. The AUC of the artificial intelligence techniques ranged from 61.0 % to 100 %, while the recall varied from 75.0 to 99.82. This research accentuates the potential of artificial intelligence techniques in the diagnosis and detection of lung cancer, with diverse levels of diagnostic accuracy. Additional research is required to optimize these artificial intelligence techniques, as well as to ascertain their clinical relevance and appropriateness in real-world clinical applicability.

Dual-degree medical students pursue additional training to prepare for careers in research, public health, and administration, but how these experiences influence residency application behaviors and outcomes are poorly understood. We analyzed 36,298 residency applicants from the Texas Seeking Transparency in Application to Residency (TexasSTAR) database spanning 2017-2023 to compare application, interview, and match patterns among single-degree MD applicants and those with MD-PhD, MD-MPH, MD-MBA, or MD-MSc degrees. Despite differences in academic metrics, application strategies, and interview rates, match rates were similar across degree groups. MD-PhD students applied to fewer programs but had the highest interview offer-to-application rate and matched at more prestigious programs based on Doximity rankings. Beyond traditional application metrics such as board scores, research productivity, grades, and honor society membership, strategies including away rotations, geographic preferencing, and program signaling were associated with increased interview offers and match success among all applicants but were less influential for dual-degree applicants. These findings suggest dual-degree applicants require specialized advising and evaluation.

To develop and evaluate an automated system for identifying healthcare barriers focusing on transportation issues in veterans' clinical notes using large language models (LLMs) and to assess the impact of different prompting strategies on classification performance and explanation consistency. We developed a hybrid system combining pattern matching for templated notes with LLM analysis for free-text notes. Using 2000 manually annotated clinical notes, we compared four prompting strategies (dual-role short, dual-role long, analysis-first, analysis-only) across Mistral-7B and Llama-3.1 models. We evaluated classification performance using standard metrics and assessed explanation consistency through embedding similarity analysis. The analysis-first strategy achieved superior performance, with Mistral-7B reaching an F1 score of 0.914, outperforming traditional machine learning approaches (GBM: 0.786, BERT: 0.811). LLMs demonstrated higher explanation consistency within models (mean cosine similarity 0.887-0.908) compared to cross-model similarities (0.767-0.872). Pattern matching successfully handled 6.7% of templated notes deterministically. Mistral-7B showed greater internal consistency but higher abstention rates compared to Llama-3.1. Requiring LLMs to analyze evidence before classification improves both accuracy and explanation consistency for identifying transportation barriers in clinical notes. This approach enables automated barrier detection at scale while providing clinically relevant explanations, supporting both population-level healthcare planning and individual patient care decisions.

This article proposes the Canon to Code (C2C) Auditing Framework for evaluating generative (artificial intelligence) AI output through classical rhetoric, arguing that AI's characteristic failures—guessing instead of knowing, politeness instead of credibility, and confidence instead of judgment—revisit problems that rhetoric has addressed since antiquity. Developed using a rulemaking methodology and drawing on classical rhetorical theory, this framework presents 10 auditing rules that operationalize rhetorical principles into evaluation criteria for AI-generated content, focusing on accuracy, transparency, and accountability. It offers content auditors, technical communicators, and compliance professionals a theoretically grounded method for distinguishing AI output that meets audience needs from output that simulates credibility through pattern matching.

Background: Canada is currently experiencing a health workforce crisis, reflective of the global workforce challenges. Integrated models of care such as Ontario Health Team (OHTs), in Ontario, Canada are important examples of large-scale reform focusing on local populations and a regional health workforce. To ensure a workforce that meets the needs of their population, a regional approach to workforce planning is required. While significant work has been published on health workforce planning, there are few examples of how regional integrated care systems can apply workforce planning models to address population needs. The overall aim of this project is to inform policy to support regional workforce planning for integrated care strategies. Approach: An exploratory mixed methods single case study design was used. The study was conducted in partnership with one Ontario Health Team that served as the case. Lived experience patient partners were core members of the research team, providing input throughout the project. The Health Workforce Planning Model was identified in a scoping review and applied to the OHT. A mixed methods approach included multiple regional, provincial, and federal data sources to describe the population, health needs, and service providers. Interviews were conducted with the OHT leaders and community members to obtain insights into workforce planning from the regional context. A document analysis of publicly available workforce planning documents across all OHTs was completed. A deliberative dialogue was held to obtain input from the OHT, provincial and federal decision makers to inform policy recommendations. Discrete analysis was conducted for each data set and the quantitative and qualitative data was merged through a process of pattern matching. Results: Using the Health Workforce Planning Model, we attempted to capture data related to service requirements (population data demographics, health status, health services utilization) and service capacity (health workforce) to examine fit and identify gaps. Data were gathered by a complex web at federal, provincial, and regional levels. Six federal/provincial data sets were used to describe the population. Nineteen of 93 regional organizations completed surveys to obtain workforce data. A review of 54 OHT websites found only 12 OHTs had documents mentioning health workforce planning, including strategic plans (66.7%) and other reports (e.g. year-end). Fourteen interviews were completed with regional OHT leaders and community partners, providing insights into contextual factors including current status of health workforce planning, driving forces, capacity for change and solutions. Key recommendations focused on the need for: 1) health workforce planning governance structures and accountability provincially and within OHTs, 2) standardized and comprehensive data and reporting across all sectors and 3) infrastructure at the provincial level and within regions to support health workforce planning. Implications: Despite the crisis in the health workforce there remains significant challenges in obtaining data to inform workforce planning and little infrastructure within integrated care systems such as Ontario Heath Teams to support regional planning. Ongoing efforts need to focus on consistency and standardization of workforce data and building in targeted resources and infrastructure to support regions to proactively plan for population needs.