This study explores the application of predictive data science techniques to anticipate customer behavior in the mobile telecommunications sector, integrating insights from finance and marketing. Leveraging advanced predictive models, including a multitask learning approach, the research is supported by an interactive web interface featuring a home- page, a Power BI dashboard, and a prediction page. The aim is to transform raw historical customer data into action- able insights for marketing teams, enabling accurate forecasts of mobile internet package activation and customers’ potential future value. Findings indicate that customer satisfaction and perceived sustainable value significantly influence subscription and recharge decisions, thereby enhancing loyalty and revenue generation. By emphasizing the synergy be- tween business needs, technological tools, and methodological frameworks, this work offers an innovative combination of theoretical and empirical approaches to advance practices within the telecommunications industry. Future research directions include incorporating real-time data streams and developing automated marketing recommendations to further optimize strategic effectiveness.

Bipolar disorder (BD) is a complex mood disorder characterized by recurrent depressive and manic/hypomanic episodes, accompanied by significant cognitive dysfunction and emotional dysregulation. Accurate and timely diagnosis, especially the differentiation between subtypes, remains a challenge due to overlapping symptoms, variable onset times for more specific symptoms (e.g., psychotic features), and the reliance on subjective assessments. This study examines the use of a multimodal approach combining electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) to identify patterns of BD emotional dysregulation, aiming to enhance its diagnosis and subtype differentiation. The protocol employed an emotional visual task to evaluate the interference of emotional content on cognitive function. EEG data were collected using a whole-head cap, while fNIRS focused on hemodynamic changes in the frontal cortex. Furthermore, the feasibility of using a potential simplified, portable EEG-fNIRS system was explored by focusing the analysis on frontal regions. The cohort included BD patients [BP] of two main subtypes, and healthy controls [HC]. Behavioral analysis revealed significant performance differences between BP and HC groups. While EEG alone enabled groups' classification, integrating EEG and fNIRS improved accuracy by reducing misclassification rates. Although classification using only frontal EEG regions was slightly less accurate than the full-head cap, fNIRS integration ensured robust results, supporting the feasibility for a potential simplified system. These findings underscore the complementary strengths of EEG and fNIRS in capturing neural and vascular markers of emotional dysregulation in BD and support the development of multimodal diagnostic tools for BD.

Premature ejaculation (PE) is one of the most common forms of male sexual dysfunction, yet its underlying neurobiological mechanisms remain unclear. This study aims to explore the role of S100 calcium-binding protein B (S100B) in PE and its regulatory relationship with brain-derived neurotrophic factor (BDNF) and serotonin (5-HT) signaling. A rat model of PE was established using behavioral screening criteria. Sexual behavior parameters were recorded, and the expression levels of S100B, BDNF, and 5-HT in brain tissues were measured using enzyme-linked immunosorbent assay, quantitative real-time PCR, Western blotting, immunohistochemistry, and immunofluorescence. The impact of S100B knockdown on PE-related behaviors and molecular expression was evaluated. The primary outcome was the effect of S100B regulation on PE-related behaviors and its interaction with the BDNF/5-HT signaling pathway. PE rats exhibited classical behavioral features, including shortened ejaculation latency and increased ejaculation frequency. Transcriptomic and protein analyses showed that S100B expression was significantly upregulated, while BDNF and 5-HT levels were markedly reduced in PE rats. S100B expression increased across several brain regions. Knockdown of S100B restored 5-HT and BDNF levels, prolonged ejaculation latency, and alleviated PE behaviors. BDNF overexpression elevated 5-HT levels and improved sexual behavior. Importantly, BDNF silencing reversed the beneficial effects of S100B knockdown, suggesting that S100B regulates ejaculation via the BDNF/5-HT pathway. Targeting S100B and its regulation of the BDNF/5-HT pathway may provide potential therapeutic strategies for managing premature ejaculation. Strengths include comprehensive molecular and behavioral analyses in a rat model provide insights into PE pathophysiology. Although this effect has been demonstrated in animal models, these models may not fully recapitulate the pathophysiological processes of human PE, and further clinical validation is required. Our findings indicate that S100B is upregulated in PE and may contribute to the pathophysiology of PE by modulating the BDNF/5-HT signaling pathway. This study provides a molecular basis for the development of therapeutic strategies targeting PE.

The µ-opioid receptor (MOP) is a critical pharmaceutical target that mediates both the therapeutic benefits and adverse effects of opioid drugs. However, the large-scale neural circuit dynamics underlying key opioid effects, such as analgesia and respiratory depression, remain poorly understood, hindering the development of safer analgesics. Here, we present a multimodal experimental framework that integrates functional ultrasound imaging through the intact skull with behavioral and molecular analyses to investigate opioid-induced large-scale functional responses and their physiological relevance in awake, behaving male mice. Administration of major opioids-morphine, fentanyl, methadone, and buprenorphine-elicited robust, dose- and time-dependent reorganization of functional brain connectivity (FC) patterns, with magnitude scaling according to MOP agonist efficacy. This opioid-specific functional fingerprint is marked by decreased FC between the somatosensory cortex and hippocampal/thalamic regions and increased bilateral FC within the somatosensory cortex. Notably, this fingerprint was attenuated following tolerance induction and abolished by pharmacological or genetic MOP inactivation. Through power Doppler spectral analysis and lagged correlation measurements, we show that morphine perturbs temporal FC dynamics and the propagation of brain-wide oscillatory activity, disrupting critical-state dynamics. Importantly, we identify a dissociation between fast, transient processes-such as cerebral blood volume changes, locomotion, and respiratory depression-and slower processes driving FC reorganization, analgesia, and sustained MOP activation. This study provides mechanistic insights into opioid-induced network reorganization, establishes FC alterations as a reliable biomarker of opioid efficacy, and offers a framework for advancing the development of analgesic compounds with improved therapeutic windows and reduced side effects.

Fast reaction is critical in sports, and understanding the factors that influence it can be a valuable contribution for coaches and sports scientists. This cross-sectional study presents a descriptive analysis of reaction time (RT) in a thousand adolescents and explores the influence of gender, age, and sports activity. Simple (SRT) and choice (CRT) reaction times are measured using a standardized electronic device. Results showed a decrease in RT with age, indicating improved neurocognitive efficiency during adolescence. No remarkable gender differences were found for either SRT or CRT. Contrary to expectations, competitive sports seem to have no impact on RT, and some sedentary individuals exhibited faster RTs than athletes. Mixed results were observed when comparing tennis players and athletes from other sports between the two genders. The outcomes suggest that RT may be more influenced by specific training and other factors (i.e., cognitive-attentive and personal) rather than general sports participation.

Abstract Sleep is an important biological function that is necessary for both mental and physical health. Assessing breathing patterns and tracking postural changes might yield important information about sleep-related behavior, particularly in people with limited mobility. In this work, a pressure mat-based system for classifying sleep posture and auxiliary respiratory monitoring through time-series analysis and deep learning is presented. For participants in supine positions, time-series analysis is used to estimate breathing rate from pressure sensor data in addition to posture recognition. To calculate the average breathing rate, respiratory-induced pressure changes are taken out of the data from the chest region, smoothed using Savitzky-Golay and moving average filters, and then examined using peak detection. A heuristic sleep behavioral stability indicator was developed by further analyzing the dynamics of posture transitions and breathing rate variability. Pressure map data from two publicly accessible datasets, PmatData and SLP, are used to classify sleep postures, taking into consideration both image-based and video-based representations. A lightweight 2D Convolutional Neural Network (2D CNN) is used for image classification, with 98.69\% accuracy on the SLP dataset and 99.84\% accuracy on the PmatData dataset. A 2D CNN is combined with recurrent architectures such as Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), and Simple Recurrent Neural Network (SimpleRNN) to create hybrid models that capture temporal dynamics for video-based categorization. The 2D CNN–SimpleRNN model performs the best among them, with low computational cost and accuracies of up to 99.15\% on PmatData and 96.18\% on SLP. All things considered, the suggested framework shows how pressure mat sensing, when combined with effective deep learning and signal processing methods, allows for precise posture categorization and low-cost supplementary respiratory monitoring. These results demonstrate the potential of non-invasive pressure-based systems for continuous sleep behaviour analysis in both clinical and residential environments.

This article examines, through a systematic literature review, the application of corporate governance practices in small and medium-sized enterprises (SMEs). The investigation focuses on scientific production from the last five years, seeking to understand the inherent limitations of traditional governance models — originally designed to meet the demands of large publicly traded corporations — and to map the adaptation alternatives being developed by the academic community. The research was conducted through a structured search in Scopus, Web of Science, SciELO, and Google Scholar databases, employing qualitative analysis of the selected works. The findings reveal that, although the adoption of formal governance mechanisms in SMEs is still in an embryonic stage, the implementation of adapted practices — notably the promotion of transparency, the establishment of accountability mechanisms, and the constitution of advisory boards — produces significant positive effects. Such practices not only expand the possibilities for raising financial resources but also enhance innovative capacity, strengthen organizational resilience, and increase the competitiveness of these companies. The conclusion points to the absence of a universal governance model applicable to SMEs, making it essential to adopt flexible and gradual approaches that consider the specificities of this business segment: ownership concentration, the lack of distinction between management and ownership, family ties, and resource constraints. Future investigations should prioritize longitudinal studies and behavioral analyses of the actors involved in governance processes.

To develop smart materials with tailored functional response, the combination of poly-(vinylidene fluoride) (PVDF) and advanced ionic additives such as ionic liquids (ILs) is increasingly being investigated. Depending on the processing conditions, the incorporation of these additives into PVDF, together with their functional response, promotes the nucleation of specific electroactive phases. This work explores the effect of incorporating sodium tetra-(2-thenoyltrifluoroacetonate) europate-(III), Na-[Eu-(tta)4] and 1-butyl-3-methylimidazolium tetra-(2-thenoyltrifluoroacetonate) europate-(III), [Bmim]-[Eu-(tta)4], into PVDF matrices through a comprehensive analysis of isothermal crystallization behavior, morphological features, crystalline phase development, and dielectric behavior. Field-emission scanning electron microscopy (FESEM) was used to analyze the microstructure, while Fourier transform infrared (FTIR) spectroscopy was used to assess the development of PVDF crystalline phases during its isothermal crystallization at various temperatures. All samples exhibited α, β, and γ crystalline phases, although their relative proportions differed significantly depending on the type of filler used. This suggests that [Bmim]-[Eu-(tta)4] is a strong promoter of the electroactive (EA) phases of PVDF. The results are attributed to the interaction between the IL charges and the PVDF dipoles of the EA structures, which are promoted by higher crystallization temperatures, as supported by both FTIR and DSC data. Thus, the addition of Na-[Eu-(tta)4] and [Bmim]-[Eu-(tta)4] strongly influences the crystallization kinetics of PVDF and allows nucleation of specific phases of PVDF. Additionally, dielectric spectroscopy revealed that the nature of the cation strongly influences conductivity behavior, as demonstrated by the dielectric results. Overall, the incorporation of Na-[Eu-(tta)4] and [Bmim]-[Eu-(tta)4] not only influences the crystallization kinetics of PVDF but also provides PVDF with intrinsic functional properties such as luminescent behavior and improved electrical performance, offering a simple and efficient strategy of nucleating specific PVDF phases.

BackgroundDepression is a global health concern, and acupuncture has emerged as an effective treatment. The role of intestinal microbiota in depression remains unclear. This study, utilizing 16S rRNA high-throughput sequencing, aimed to explore the relationship between electroacupuncture (EA) and depressive behavior by examining changes in the intestinal microbiota.Materials and MethodsForty-eight male Sprague-Dawley rats were utilized, with 13 assigned to the normal control (NC) group. The remaining rats underwent a 28-day depression modeling process, and those exhibiting depressive symptoms were randomly divided into chronic unpredictable mild stress (CUMS) and EA groups. The EA group received 14 days of treatment. Behavioral analyses were conducted on rats from the NC, CUMS, and EA groups to assess EA’s effectiveness. Additionally, 16S rRNA sequencing was performed on randomly selected rats from each group.ResultsBy examining the behavior of 39 rats and the intestinal microbiota of 18 rats, we found that EA may alter the composition of the intestinal microbiota community structure in CUMS rats, particularly modulating the abundance of Akkermansia in a manner potentially linked to gut–brain axis regulation, including stress- and inflammation-related pathways that may influence microbial composition.ConclusionThe potential antidepressant impact of electroacupuncture (EA) might be linked to the modulation of Akkermansia abundance within the brain-gut axis.

Global financial integrity is fundamentally challenged by cryptocurrency mixers such as Tornado Cash, which facilitate billions in illicit fund flows. Low detection rates, reliance on labeled training data that is unavailable for novel attacks, and failure to analyze temporal coordination patterns are all impediments to the effectiveness of existing forensic tools. We introduce CONSENSUS, a self-supervised heterogeneous ensemble framework that addresses the challenge of attribution in mixed transaction streams. Our system requires no pre-existing labels, and it generates supervision signals directly from on-chain behavioral patterns. It synthesizes evidence by orchestrating nine analytical modalities—including deterministic clustering, behavioral analysis, and multiple graph neural network architectures—through a formal consensus mechanism. This multi-modal approach produces transparent, auditable risk scores from a 111-dimensional behavioral fingerprint. We validated the framework on five major decentralized finance (DeFi) exploits, including the Ronin Bridge and Poly Network hacks. Using raw transaction data, it detected all known primary attackers at 100% accuracy without training. Crucially, the framework's self-supervised components successfully identified the novel attack pattern of the Poly Network exploit, thereby demonstrating robustness to out-of-distribution threats that defeat supervised methods. By providing a transparent, zero-label solution, CONSENSUS establishes a new paradigm for flexible, effective risk profiling and forensic investigation.

Behavioral and Questionnaire-Based Analysis of Auditory Processing Disorder in Korean Patients

Abstract A fundamental understanding of landslide evolution requires characterizing how deformation localizes within the sliding mass, as these non-homogeneous zones provide crucial insights into how destabilization initiates, failure surfaces develop, and the overall kinematic behavior evolves. While traditional analysis often assumes uniform movement, this study presents a methodology to quantify intricate patterns of surface deformation at a fine scale, allowing for the direct analysis of localization behavior. By applying strain tensor analysis to high-resolution displacement fields derived from multi-temporal Uncrewed Aerial Vehicle-Light Detection and Ranging (UAV-lidar) and Structure from Motion (SfM) surveys, we compute the divergence, gradient, and curl fields for two distinct landslides: one translational and one rotational. This approach quantifies volumetric changes, translational strain, and rotational components, revealing unique kinematic signatures for each landslide type. The translational slide is characterized by alternating expansion-contraction patterns along its dip-line, whereas the rotational slide exhibits clear, separate bands of head subsidence and toe expansion, coupled with non-uniform rotation along the strike. This detailed characterization of strain localization provides direct observational evidence of the fundamental mechanisms governing landslide behavior. It offers a more nuanced, mechanistic understanding that advances the interpretation of slope instability, providing a stronger physical basis for hazard assessment and risk management.

Artificial intelligence (AI) provides new opportunities for high-resolution behavioral analysis and automated, human-free experiments. Here we present AVATAR (AI Vision Analysis for Three-dimensional Action in Real-time). This AI-driven system reconstructs 3D mouse motions by detecting key body parts from synchronized multi-view videos and converting into action skeletons. AVATAR achieves near-human accuracy in pose estimation, enables robust extraction of kinematic and postural features, and supports scalable analysis of model animal behaviors. Using these features represented by 3D action skeleton, LSTM-based model reliably classifies freely moving mouse behaviors during various experimental paradigms with low-latency processing (100 ms) enables real-time closed-loop optogenetic stimulation. As a demonstration of generalizability, we applied AVATAR framework to bottom-view predatory hunting paradigm. AVATARnet accurately detected mouse poses and extracted dynamic behavioral features of the mouse. Using AVATARnet-driven dynamic features, an XGBoost-based classifier automated action segmentation annotation during complex predatory chasing behavior. Together, AVATAR provides 3D pose estimation, dynamic quantification, classification, and closed-loop manipulation in real-time.

Abstract In recent decades, the extent of ice-free areas has been increasing in the South Shetland Islands (Maritime Antarctica). The coastal sector is one of the zones most significantly affected by glacial retreat, with newly exposed land surfaces undergoing a wide variety of post-glacial environmental processes. Coastal areas are characterized by both continental and marine ice dynamics, which in turn have major influences on the morphology and processes shaping coastal landforms. A detailed geomorphological analysis was carried out at Spanish Cove, south-west Livingston Island, which constitutes a boulder beach close to the Spanish Antarctic Station Juan Carlos I. This research provides a classification of the existing coastal landforms in this sector, as well as an analysis of the recent behaviour of the area using drone surveys, material size measurements obtained through semi-automatic techniques and hardness analysis using a durometer. This study represents one of the first attempts to classify the Antarctic coastal environment and offers a basis for understanding the potential evolution of such environments over the coming decades under global change and the rapid transformation of present-day glaciated landscapes.

Thermal behavior and kinetic analysis of lotus leaf by non-isothermal procedures

Computational analysis of therapeutic communication presents challenges in multi-label classification, severe class imbalance, and heterogeneous multimodal data integration. We introduce a bidirectional analytical framework addressing patient emotion recognition and provider behavior analysis. For patient-side analysis, we employ ClinicalBERT on human-annotated CounselChat (1482 interactions, 25 categories, imbalance 60:1), achieving a macro-F1 of 0.74 through class weighting and threshold optimization, representing a six-fold improvement over naive baselines and 6–13 point improvement over modern imbalance methods. For provider-side analysis, we process 330 YouTube therapy sessions through automated pipelines (speaker diarization, automatic speech recognition, temporal segmentation), yielding 14,086 annotated segments. Our architecture combines DeBERTa-v3-base with WavLM-base-plus through cross-modal attention mechanisms adapted from multimodal Transformer frameworks. On controlled human-annotated HOPE data (178 sessions, 12,500 utterances), the model achieves a macro-F1 of 0.91 with Cohen’s kappa of 0.87, comparable to inter-rater reliability reported in psychotherapy process research. On YouTube data, a macro-F1 of 0.71 demonstrates feasibility while highlighting annotation quality impacts. Cross-dataset transfer and systematic attention analyses validate domain-specific effectiveness and interpretability.

Objective: The aim of this study was to investigate the facial expressions of children diagnosed with attention deficit hyperactivity disorder (ADHD) using computerized facial analysis and to examine their emotion recognition abilities. Methods: A total of 56 children with ADHD and 45 control subjects aged 6-12 years were included. The Diagnostic Analysis of Nonverbal Expressions-2 (DANVA) was used to measure the participants’ emotion recognition abilities. One group of participants watched animated film scenes lasting an average of 7 minutes, and their facial behaviors were recorded on video. OpenFace software was used for video analysis. Support Vector Machines (SVM), naive Bayes, and logistic regression machine learning methods were used to distinguish between the data of the ADHD and control groups. Results: The significant difference found in DANVA total scores indicating poorer emotion recognition skills in ADHD was not significant when intelligence levels were controlled. Children with inattention as the primary symptom made significantly more errors in emotion recognition from posture and total scores in DANVA child faces and overall compared to the other groups. According to computerized facial analysis, Video 1, which predominantly featured fear and anger emotions, was the most distinctive video for both healthy controls and the ADHD group. When analyzing AU units, AU12 (lip corner pulling), AU07 (eyelid raising), AU09 (nose wrinkling), AU45 (eye blinking), and AU06 (cheek raising) were the most distinctive features. Conclusion: Emotion recognition levels differed among ADHD cases according to clinical subtypes and comorbid psychiatric disorders. The most significant difference between the ADHD and control groups during emotion-containing video viewing was observed while watching sad videos. The findings of this study can be considered promising for the diagnostic validity of machine learning methods in ADHD, oneof the most common neurodevelopmental disorders. Cite this article as: Karabekiroglu K, Usta MB, Kesim N, Şahin İ, Ayyıldız M. Computerized analysis of face emotion recognition skills and facial behaviors in children with attention deficit hyperactivity disorder. Neuropsychiatr Invest. 2026, 64, 0054, doi:10.5152/NeuropsychiatricInvest.2026.25054.

The vestibular system is a critical sensory modality required for coordinated movement, balance and our ability to interact with the surrounding environment. Vestibular sensory neurons provide the nervous system with information about head rotation and acceleration. However, the nervous system can also modify the activity of sensory neurons and hair cells via the actions of the efferent vestibular system (EVS). The function of the EVS has remained unknown partly because of an inability to target efferent vestibular neurons in a selective manner to understand their synaptic inputs and function during behaviour. Here, we present a novel method for the selective targeting and expression of flp-recombinase in EVS neurons. We take advantage of the dual expression of choline acetyl transferase (ChAT) and calcitonin gene related peptide (CGRP) in these neurons to develop an adeno-associated virus (AAV) that expresses a gene only in neurons with this intersectional expression. We use this system to map the monosynaptic inputs to EVS neurons and show inputs from distinct populations of brainstem and midbrain regions indicating a functional role as a multimodal processing center and integrator for the vestibular periphery. To demonstrate the applicability of our technology in behavioural assays, we performed a preliminary behaviour analysis (treadmill running and open field) in mice with disrupted EVS function. While more bespoke assays are required to ascertain EVS function/s, our viral method presents a novel tool for investigators examining the role of the vestibular system and its central circuits.

Global functional connectivity of cognitive control networks predicts task-switching performance in older adults.

Analysis of motorcycle rider behavior and attitudes in urban and rural environments: comparative study utilizing a motorcycle rider behavior questionnaire