Kocaeli University Research Information System
Not a member yet
80615 research outputs found
Sort by
From LSTM to GPT-2: Recurrent and Transformer-Based Deep Learning Architectures for Multivariate High-Liquidity Cryptocurrency Price Forecasting
No full textThis study introduces a unified and methodologically symmetric comparative framework for multivariate cryptocurrency forecasting, addressing long-standing inconsistencies in prior research where model families, feature sets, and preprocessing pipelines differ across studies. Under an identical and rigorously controlled experimental setup, we benchmark six deep learning architectures—LSTM, GPT-2, Informer, Autoformer, Temporal Fusion Transformer (TFT), and a Vanilla Transformer—together with four widely used econometric models (ARIMA, VAR, GARCH, and a Random Walk baseline). All models are evaluated using a shared multivariate feature space composed of more than forty technical indicators, identical normalization procedures, harmonized sliding-window formations, and aligned temporal splits across five high-liquidity assets (BTC, ETH, XRP, XLM, and SOL). The experimental results show that transformer-based architectures consistently outperform both the recurrent baseline and classical econometric models across all assets. This superiority arises from the ability of attention mechanisms to capture long-range temporal dependencies and adaptively weight informative time steps, whereas recurrent models suffer from vanishing-gradient limitations and restricted effective memory. The best-performing deep learning models achieve MAPE values of 0.0289 (BTC, GPT-2), 0.0198 (ETH, Autoformer), 0.0418 (XRP, Informer), 0.0469 (XLM, Informer), and 0.0578 (SOL, TFT), substantially improving upon the performance of both LSTM and all econometric baselines. These findings highlight the effectiveness of attention-based architectures in modeling volatility-driven nonlinear dynamics and establish a reproducible, symmetry-preserving benchmark for future research in deep-learning-based financial forecasting
MONITORING STUDENTS' LOGICAL THINKING SKILLS IN AN INTERVENTION PROGRAM USING LATENT GROWTH MODELS
No full textComparative thermal analysis of heat transfer fluid injection orientation in phase change material based latent heat thermal energy storage systems
No full textThis study aims to experimentally investigate the influence of heat exchanger orientation and heat transfer fluid (HTF) injection direction on the thermal performance of a latent heat thermal energy storage (LHTES) system using a phase change material (PCM). A comparative analysis is conducted between horizontal and vertical double-pipe heat exchanger configurations, considering both top and bottom injection of the HTF. The experimental methodology involves systematic variations of inlet temperature and volumetric flow rate while monitoring temperature evolution, charging and discharging durations, heat transfer coefficients, and Nusselt numbers. The results demonstrate that the direction of HTF injection plays a critical role in governing heat transfer behavior. Top injection promotes faster charging by enabling the PCM to reach higher average temperatures, whereas bottom injection enhances convective heat transfer, yielding the highest Nusselt numbers and heat transfer coefficients, albeit with longer charging and discharging times. In the horizontal configuration, increasing the flow rate from 60 to 120 L h−1 reduces the charging time from 70 to 60 min, corresponding to an improvement of 14.2%. Discharge tests performed at inlet temperatures of 70 °C and 80 °C reveal performance reductions of 8.9% for the horizontal configuration, 43.3% for bottom injection, and 42.6% for top injection at a constant flow rate. Overall, the horizontal configuration exhibits the fastest thermal response during both charging and discharging processes, while bottom injection achieves the highest thermal efficiency. These findings highlight the importance of heat exchanger orientation and flow direction optimization for improving the performance of LHTES systems
Mesoporous Black Arsenic: Wet-Chemical Synthesis and Its Photo-Enhanced Activity in the C-H Arylation of Heteroarenes under Red Light Illumination
No full textBlack arsenic (b-As) is theoretically known for its anisotropic electronic and optical properties, yet its experimental realization remains limited. Here, we report the first bottom-up wet-chemical approach for the synthesis of mesoporous b-As (mb-As), a structurally ordered As-As networked semiconducting material. The as-synthesized mb-As exhibits a direct bandgap of 1.75 eV and a conduction band minimum of -1.40 V vs normal hydrogen electrode, providing a stronger reduction potential than most conventional semiconductors and enabling efficient visible-light-driven single-electron transfer (SET) under 625 nm illumination. These properties were leveraged in photoredox C-H arylation of heteroarenes, where mb-As catalyzed 13 transformations with yields up to 88%, successfully accommodating substrates bearing electron-donating and withdrawing groups as well as aromatic heterocycles. Mechanistic investigations revealed a radical-driven SET pathway involving both photogenerated electrons and holes of mb-As. Importantly, to address concerns regarding the potential toxicity of mb-As, biocompatibility assays performed on two different cell lines demonstrated that >90% cell viability up to 1 mg/L and ca. 70% cell viability at 50 mg/L, distinguishing mb-As from highly toxic arsenic analogues such as As(III) chloride and underscoring its suitability for sustainable catalysis and other possible applications
Systemic immune–inflammation index dynamics during methotrexate therapy for psoriasis: a retrospective cohort study
No full textWhile methotrexate (MTX) modulates systemic inflammation in psoriasis, its effects on novel hematologic indices and their clinical utility remain underexplored. To investigate MTX-induced changes in systemic inflammatory markers and their clinical correlations in psoriasis. This retrospective study analyzed 70 adults with moderate-to-severe psoriasis receiving subcutaneous MTX (15-25 mg/week) for 3 months. We evaluated changes in CBC-derived indices including neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), systemic immune-inflammation index (SII), and systemic inflammation response index (SIRI), acute-phase reactants C-reactive protein(CRP), erythrocyte sedimentation rate(ESR), and their associations with PASI75 response. Subgroup analyses assessed psoriatic arthritis and regional involvement (scalp, nail, genital) effects. Primary endpoint was the percent change in the systemic immune-inflammation index (SII) from baseline to Month 3; secondary endpoints included percent changes in NLR, PLR, SIRI, ESR, and CRP, and their correlations with clinical response (PASI and PASI75). At Month 3, PASI decreased from median 7.45 (IQR 4.80–10.82) to 2.75 (IQR 1.20-5.28) (p<0.001). NLR (p=0.029), SII (p=0.031), ESR (p<0.001), and CRP (p=0.004) declined. PASI% change correlated weakly with SII% change (ρ=0.255, p=0.033). Compared with non-responders, PASI75 responders showed a greater SII reduction (−9.89% vs +7.38%; p=0.041). Genital involvement was associated with pronounced decreases in NLR (−21.0%, p=0.030), SII (−24.1%, p=0.016), and PLR (−23.3%, p=0.001), while psoriatic arthritis attenuated ESR reduction (p=0.049). This study identifies SII as a responsive hematologic marker of MTX efficacy in psoriasis, correlating with clinical improvement and showing greater reductions in treatment responders. The differential inflammatory marker changes observed across disease phenotypes highlight the potential for tailored monitoring approaches using routine CBC indices
Effectiveness of mobile-based monitoring system (ONKOSIS) in the management of chemotherapy-related symptoms: a randomized controlled trial
No full textPurpose: This study was conducted to evaluate the impact of the ONKOSIS mobile application, developed within the scope of the study, on the management of chemotherapy-related symptoms and quality of life. Methods: This randomized controlled, single-blind experimental design study was carried out with the participation of 199 patients at the Kocaeli University Semahat Aracı Oncology and Palliative Care Center, Outpatient Chemotherapy Unit. In the study, which used simple randomization, the intervention group consisted of 98 patients and the control group consisted of 101 patients. During chemotherapy treatment, in addition to standard care for symptom management, the intervention group used the ONKOSIS mobile application, while the control group used a demo mobile application. The study data were obtained through the Diagnostic Form, Nightingale Symptom Assessment Scale, Coping with Chemotherapy Symptoms Form, and patient satisfaction survey. The follow-up frequency was planned according to the type of treatment protocol applied to the patients. Patients who received four cycles of treatment were followed up after the 2nd, 3rd, and 4th cycles, while patients who received six cycles of treatment were followed up after the 2nd, 4th, and 6th cycles. Results: The study group had a mean age of 47.18 ± 13.83; 74.4% were female. Treatment cycles of 2, 3, or 4 weeks were provided for 26.6% of patients receiving treatment for gastrointestinal cancer, 21.6% for breast cancer, and 17.1% for lung cancer. Sociodemographic and clinical characteristics were similar between groups. Although the intervention group exhibited greater initial symptom burden (higher N-SAS scores at T1/T2), their symptom trajectory improved significantly over time compared to the control group, supporting the role of mHealth in enhancing outcomes. The use of ONKOSIS improved quality of life and patient satisfaction but did not reduce unscheduled healthcare visits. Conclusion: The ONKOSIS mobile application was found to be an effective tool in supporting symptom management and improving the quality of life for cancer patients during the treatment process. It was determined that the widespread adoption of the mobile application in cancer care could contribute to improving patient outcomes and enhancing care quality. Trial registration: ClinicalTrials.gov ID: NCT05946070. Registered on June 22, 202
The Effect of Pain Catastrophizing and Pain Intensity on Postoperative Mobility in Patients Undergoing Laparoscopic Abdominal Surgery: A Prospective Cross-sectional Correlational Study
No full textPurpose This study examines how pain catastrophizing and pain intensity influence postoperative mobility in patients undergoing laparoscopic abdominal surgery. Design The study has a prospective cross-sectional correlational design. Methods The study included 92 patients who underwent laparoscopic abdominal surgery. Data were collected using the Patient Information Form, Pain Catastrophizing Scale (PCS), Numeric Rating Scale, Patient Mobility Scale (PMS), and Observer Mobility Scale (OMS). Preoperative pain catastrophizing was assessed. Patients whose vital signs and pain intensity were evaluated during any mobilization after the first mobilization within 48 hours postoperatively were mobilized. The level of pain, difficulty, and dependence experienced during physical movement was assessed. Findings The mean age of the patients was 56.12 ± 13.26 years, 54.3% were female, 66.3% underwent laparoscopic cholecystectomy, 19.6% underwent laparoscopic hernia repair, and 14.1% underwent laparoscopic appendectomy. Pain intensity, PCS score, PMS score, OMS score, and body mass index (BMI) were found to be positively correlated ( P < .05). After mobilization, there was a significant increase in pulse rate, respiratory rate, mean blood pressure, and pain intensity within the range of clinically normal parameters ( P < .01). Medical devices (urinary catheter, drain) were found to cause an increase in PCS, PMS, OMS, and pain severity. In the multivariate regression model, pain intensity, pain catastrophizing, and BMI were found to explain 56% of patient mobility ( F (4, 87) = 29.896, P < .01). Conclusions The study results showed that pain, pain catastrophizing, and BMI negatively affected postoperative patient mobility. These barriers need to be taken into consideration when making interventions to improve patient mobility
CFD analysis on thermo-hydraulic performance behavior of an indirect solar dryer containing various configurations of rectangular ribs
No full textSolar drying systems often suffer from low thermal efficiency due to poor heat transfer in the solar air heater (SAH). To address this limitation, this study numerically investigates the effect of incorporating rectangular ribs, both perforated and non-perforated ribs, placed beneath the absorber plate of a solar air heater used in an indirect solar dryer. 2D-CFD simulations were performed using ANSYS Fluent 17.1, employing the RNG k–ε turbulence model with enhanced wall treatment to accurately resolve near-wall flow behavior. The analysis considered a constant heat flux of 1000 W m−2, Reynolds numbers between 4000 and 18,000, a rib blockage height ratio of 0.23–0.7, and a rib spacing ratio of 3.33–10. Results showed that the incorporation of rectangular ribs as roughness elements significantly improves the heat transfer and friction characteristics. The average Nusselt number increased by up to 2.05 times for non-perforated ribs and 1.89 times for perforated ribs compared with a smooth duct. However, this enhancement was accompanied by an increase in the friction factor of 64.54 times and 5.17 times, respectively. The optimal thermo-hydraulic performance parameter of 1.12 was achieved for perforated ribs at BR = 0.7, PR = 3.33, and Re = 4000. Additionally, roughened configurations raised the drying air temperature by 13.25 K for non-perforated ribs and 5.62 K for perforated ribs, improving the drying chamber heat gain by 39.38 and 16.33%, respectively. Upgrading to a roughened SAH, particularly the perforated ribs configuration, significantly enhances heat transfer, airflow mixing, and energy utilization. This improvement offers a practical and efficient approach to increasing the thermal performance of solar drying systems
Empowering Mothers: The Effects of Audio-Guided Meditation on Resilience and Spiritual Well-Being of Mothers of Children Diagnosed with Cancer in Türkiye—A Randomized Controlled Trial
No full textThe diagnosis of childhood cancer profoundly affects family life, and mothers who typically assume the primary caregiving role are particularly vulnerable to reduced resilience and diminished spiritual well-being. This randomized controlled trial aimed to evaluate the effect of audio-guided meditation on resilience and spiritual well-being among mothers of children aged 6–18 years with oncology diagnoses. The study was conducted between April and August 2024 with 60 mothers in a pediatric hematology–oncology clinic in Western Türkiye. Data were collected using the Descriptive Information Form, the Connor–Davidson Resilience Scale–Short Form (CD-RISC), and the Spiritual Well-Being Scale (SWBS). The audio-guided meditation group participated in four weeks of daily audio-guided meditation, while the control group received no intervention. Most participants were aged 26–35 years (33.0%), half were high school graduates (50%), and 47.5% had another child. Leukemia was the most common diagnosis (30.0%), and 51.7% of children had been hospitalized for more than six months. The audio-guided meditation group showed significant increases in mean CD-RISC and SWBS scores at posttest compared with baseline (p 0.05). Significant improvements were also noted in the SWBS subscales of Transcendence, Harmony with Nature, and Anomie, with posttest Anomie scores differing significantly between groups (p = 0.010). These findings suggest that audio-guided meditation may be a feasible, low-cost nursing intervention to enhance resilience and spiritual well-being among mothers of children with oncology diagnoses in pediatric hematology–oncology settings. Integrating brief, structured audio-guided meditation into routine psychosocial care may help nurses support mothers’ coping resources and holistic well-being. Further research is needed to confirm its effectiveness across different clinical settings and caregiving contexts. Several limitations should be acknowledged regarding this trial. Trial registration: The clinical trial was registered (ClinicalTrials.gov: NCT06484283). Registered on 7th June 2024
On multiplicative beta-convexity and associated Katugampola fractional inequalities: Hermite–Hadamard and parametrized Newton–Cotes results
No full textIn this work, we introduce a new class of functions, referred to as multiplicative beta-convex functions, which extends several existing convexity concepts in the context of multiplicative calculus. Based on this notion, we establish a Hermite–Hadamard inequality involving multiplicative Katugampola fractional integrals. We also derive a new integral identity associated with these operators, which allows us to construct a unified family of parameterized three-point Newton–Cotes-type inequalities. These inequalities apply to functions whose first-order ∗-derivatives, in ∗-absolute value, are multiplicative beta-convex. By suitably selecting the parameters, our general framework yields various well-known inequalities as special cases, including midpoint, trapezoidal, Simpson-, and Bullen-type inequalities. These arise naturally under specific subclasses of beta-convexity, such as multiplicative s-, P-, and tgs-convexity. To corroborate our theoretical results, we present a numerical example together with graphical illustrations that demonstrate the validity and sharpness of the obtained bounds. Our contributions advance the development of fractional multiplicative analysis and offer new perspectives for applications in numerical approximation and the theory of inequalities