Analysis Model Research Articles

A Novel Multi-modal Sentiment Analysis Based on Multiple Kernel Learning with Margin-Dimension Constraint

Most of the sentiment analysis studies focus on the sentimental classification of pictures in the video, ignoring the spatio-temporal information of the sequence of picture frames as well as text and audio information. The multiple kernel learning is a new hotspot in the field of nuclear machine learning, capable of handling multiple modalities. For multiple kernel learning, it is easy to ignore the basic features that are not discriminative, and cannot make full use of the base features of different modes. This paper puts forward a novel multi-modal fusion model for sentiment analysis, in which a multiple kernel learning algorithm based on convolution margin-dimension constraint is proposed for feature fusion. Moreover, the 3D convolutional neural network is used to extract the features of visual information, and the multiple kernel learning algorithm based on margin-dimension constraint is used to fuse visual, text and audio sentiment features. Experiments conducted on the MOUD and IEMOCAP sentiment databases show that the proposed model outperforms existing models in the field of multi-modal sentiment analysis research.

Exploring genotype by environment interaction in sunflower using genotype plus genotype by environment interaction (GGE) and best linear unbiased prediction (BLUP) approaches

Sunflower is important for food, feed and ornamentals, faces rising demand for quality oil. Identifying resilient high-yielding genotypes is essential. Multi-environmental trials and integrated approaches to evaluate genotype by environment interactions (GEI) enhance accuracy and comprehensiveness of breeding programs. This study assess the representativeness and discriminative power of tested environments; classify mega-environments for genotype recommendations and evaluate how GGE and BLUP methods identify stable and superior sunflower genotypes. The experiment used a randomized complete block design with three replications across six locations for two seasons. The pooled analysis of variance for GEI on seed yield showed significant variability (p < 0.05) with mean values from 1171 to 2564 kg ha−1, indicating inconsistent genotypes performance across environments. The study identified E11, E2, and E6 as key discriminating and representative environments and grouped test locations into four mega-environments. Genotypes G6, G1, and G4 are the most desirable based on stability and seed yield. E11 and G6 are identified as ideal environment and genotype, respectively. GGE biplot and BLUP methods consistently identify stable and ideal sunflower genotypes. However, GGE stability indices differ somewhat in assessing genotypes stability possibly due to low number of genotypes, testing environment, and year in the dataset. The study confirmed that simultaneous consideration of GGE and BLUP models for GEI analysis helps to select more stable and adaptable sunflower genotypes across environments. It also clarifies GEI pattern, identifying mega-environments, and aiding genotype selection in sunflower production. However, validating these mega-environments and exploring the genetic basis of adaptability and stability remains essential.

Performance Evaluation of Different Clustering Techniques and Parameters of Hybrid PSO- and GA-ANFIS on Optimization and Prediction of Biomethane Yield of Alkali-Pretreated Groundnut Shells

AbstractThe study focuses on optimizing biomethane yield in the anaerobic digestion of alkali-pretreated groundnut shells, involving varied input parameters. Biomethane optimization will improve the economy of the technology, which will assist in managing the environmental challenges of fossil fuel combustion. Traditional methods prove challenging, inaccurate, and uneconomical, necessitating efficient optimization models. This research hybridizes particle swarm optimization (PSO) and genetic algorithms (GA) with adaptive neuro-fuzzy inference system (ANFIS) models, assessing input parameters’ influence on biomethane yield through renowned performance metrics. Comparing the best model in the hybrid analysis, encompassing pretreatments A-E, the PSO-ANFIS (RMSE = 1.1719, MADE = 0.6525, MAE = 0.9314, Theil’s U = 0.1844, and SD = 0.7737) outperformed the GA-ANFIS (RMSE = 1.9338, MADE = 0.9318, MAE = 1.6557, Theil’s U = 0.2734, SD = 1.0598), using the same cluster radius of 0.50. Furthermore, compared to the GA-ANFIS model, the PSO-ANFIS model demonstrated significant improvements across various metrics: RMSE by 39.40%, MADE by 29.97%, MAE by 43.75%, Theil’s U by 32.56%, and SD by 27.00%. Results indicate that the PSO-ANFIS model outperforms the GA-ANFIS model, emphasizing the importance of suitable clustering algorithms and precise parameter adjustment for optimal performance in predicting biomethane yield from pretreated lignocellulose feedstocks. Graphical Abstract

Cultural adaptation and validation of the knee injury and osteoarthritis outcome score into Sinhala language in patients with primary knee osteoarthritis: a cross-sectional study

BackgroundKnee Injury and Osteoarthritis Outcome Score (KOOS) scale is a patient-reported outcome measurement tool. It evaluates both short- and long-term consequences of knee injury and primary osteoarthritis. This study aims to translate and validate the KOOS scale for a Sinhala-speaking Sri Lankan population.MethodsA cross sectional study was conducted in three hospitals. Four hundred and fifteen patients comprising 185 males and 227 females (3 subjects did not reveal their gender) with knee osteoarthritis (KOA) participated in the study. Seventy nine participants without KOA were recruited as controls. The functionality and quality of life level in patients and healthy participants were assessed using translated versions of the KOOS and Short Form-36 (SF-36) scales. Internal consistency of the instrument was assessed by Cronbach alpha. Construct validity and test-retest reliability were examined using the Intraclass Correlation Coefficient (ICC). Confirmatory Factor Analysis (CFA) was used to assess factorial validity.ResultsThe mean age (± sd) of the KOA subjects was 54.9 (± 9.2) years and for the control group was 49.2 (± 8.0) years. Majority of the respondents were female and Sinhalese in both groups. Internal consistency reliability was high (Cronbach’s alpha values ≥ 0.70). The test–retest reliability was excellent with the intraclass correlation coefficient for all subscales being above 0.90. Construct validity was assessed by the magnitude of the correlation coefficient between KOOS and SF-36 subscale scores. KOOS Pain scale moderately correlated with SF-36 bodily pain (Pearson’s r = 0.41). SF-36 physical function scores had a weak positive correlation with all KOOS subscales and SF-36 emotional wellbeing was not significantly correlated with KOOS Quality of Life (QoL) subscale. A five-factor Confirmatory Factor Analysis (CFA) model yielded a Comparative Fit Index (CFI) = 0.950, Tucker Lewis Index (TLI) = 0.946, Root Mean Square Error of Approximation (RMSEA) = 0.082 and Standardised Root Mean squared Residual (SRMR) = 0.072.ConclusionThe Sinhala translation of the KOOS scale is a reliable and valid instrument to assess KOA in a Sinhala-speaking Sri Lankan population. Studies to assess its use as a scale to evaluate responsiveness are recommended.

Dynamic response and blast resistance of I‐shaped steel‐concrete composite beam under explosive loading

AbstractTo investigate the dynamic mechanical response and damage mechanisms of I‐shaped steel‐concrete composite beams under explosive loads, experimental research and numerical simulations were conducted on steel‐concrete composite structures. The accuracy of the numerical analysis model was validated by comparing the damage characteristics of the structures obtained from explosive tests on the steel‐concrete composite components. Based on this, numerical simulations were performed on steel‐concrete composite beams using the explicit dynamic analysis software ANSYS/LS‐DYNA, and a study on their damage mechanisms was conducted. This study resulted in the acquisition of dynamic mechanical response patterns, including time‐dependent stress, strain, displacement, acceleration, and so on. Different structural damage characteristics under various explosive conditions were summarized, and key parameters affecting the blast resistance of the structure were analyzed. The research findings indicated that, in contrast to the failure characteristics of I‐shaped reinforced concrete (RC) beams, the failure characteristics of I‐shaped steel‐concrete composite beams mainly include punching and shearing failure of the steel‐RC slab and local buckling of the steel beam. Under the same explosive conditions, steel‐concrete composite structures exhibit superior blast resistance, with certain changes in structural parameters significantly improving blast resistance. The research results can provide theoretical support and a scientific basis for the proactive design of blast protection in steel‐concrete beams.

Predicting fatigue failure in five‐axis machined ball‐end milled components through FKM local stress approach

AbstractComponents created with five‐axis machining show a multi‐scale surface character due to cusps created on the surface and feed and tool marks within the cusps. Therefore, it becomes difficult to incorporate the effects of surface character on fatigue life for such components. In this work, an Forschungskuratorium Maschinenbau (FKM) guideline is adapted to develop a fatigue prediction model which considers cusps as notches and marks within the cusps as surface roughness (characterized by parameter R10z). The assessment uses stresses obtained from an finite element analysis model to predict the fatigue life of components whilst considering stress concentration, stress gradient, mean stress, and surface roughness effects. When cusps are regarded as surface roughness within the conventional FKM approach, fatigue life is considerably underestimated. In comparison, fatigue life predictions that take into consideration the roughness within cusps and treat cusps as stress‐raising notches are closer to experimental life.

Quantitative Analysis of Pb in Soil Using Laser-Induced Breakdown Spectroscopy Based on Signal Enhancement of Conductive Materials

Studying efficient and accurate soil heavy-metal detection technology is of great significance to establishing a modern system for monitoring soil pollution, early warning and risk assessment, which contributes to the continuous improvement of soil quality and the assurance of food safety. Laser-induced breakdown spectroscopy (LIBS) is considered to be an emerging and effective tool for heavy-metal detection, compared with traditional detection technologies. Limited by the soil matrix effect, the LIBS signal of target elements for soil heavy-metal detection is prone to interference, thereby compromising the accuracy of quantitative detection. Thus, a series of signal-enhancement methods are investigated. This study aims to explore the effect of conductive materials of NaCl and graphite on the quantitative detection of lead (Pb) in soil using LIBS, seeking to find a reliable signal-enhancement method of LIBS for the determination of soil heavy-metal elements. The impact of the addition amount of NaCl and graphite on spectral intensity and parameters, including the signal-to-background ratio (SBR), signal-to-noise ratio (SNR), and relative standard deviation (RSD), were investigated, and the mechanism of signal enhancement by NaCl and graphite based on the analysis of the three-dimensional profile data of ablation craters and plasma parameters (plasmatemperature and electron density) were explored. Univariate and multivariate quantitative analysis models including partial least-squares regression (PLSR), least-squares support vector machine (LS-SVM), and extreme learning machine (ELM) were developed for the quantitative detection of Pb in soil with the optimal amount of NaCl and graphite, and the performance of the models was further compared. The PLSR model with the optimal amount of graphite obtained the best prediction performance, with an Rp that reached 0.994. In addition, among the three spectral lines of Pb, the univariate model of Pb I 405.78 nm showed the best prediction performance, with an Rp of 0.984 and the lowest LOD of 26.142 mg/kg. The overall results indicated that the LIBS signal-enhancement method based on conductive materials combined with appropriate chemometric methods could be a potential tool for the accurate quantitative detection of Pb in soil and could provide a reference for environmental monitoring.

A Prediction Method for the Average Winding Temperature of a Transformer Based on the Fully Connected Neural Network

The average winding temperature of a transformer (AWTT), serving as a key indicator for assessing the running state of the transformer, is of utmost importance in determining a transformer’s electrical properties and the insulation longevity of the transformer. An accurate prediction of AWTT is essential for ensuring the safe operation of the transformer. A novel method for predicting AWTT is introduced based on the analysis of field monitoring data. Firstly, the thermal characteristics and operational mechanisms of oil-immersed transformers are examined. Secondly, a factor analysis model is developed to streamline the network structure, accounting for the strong correlations among ambient temperature, load current, and top oil temperature. Thirdly, the independent temperature factor and load factor are extracted as pivotal features, and then input into the fully connected neural network to predict AWTT. Through a case study involving a 110 kV/10 kV oil-immersed transformer, the results show that the proposed method reduces redundant correlation information compared to traditional methods and improves the prediction accuracy of AWTT, establishing a foundation for further transformer state assessments.

Exploring Weak Magnetic Signal Characteristics of Pipeline Welds: Insights into Stress Non-Uniformity Effects

Weak magnetic detection technology can detect stress concentration areas in ferromagnetic materials. However, the stress non-uniform characteristics of pipeline welds lead to significant differences in stress distribution range and values between inner wall welds and outer wall welds. This discrepancy makes it crucial to further evaluate the impact of stress non-uniformity on magnetic signals. To study the magnetic signal characteristics under the influence of residual stress in weld seams, a magneto-mechanical analytical model was established based on the magnetic charge theory and the distribution characteristics of residual stress in the weld seam. The magneto-mechanical relationship and magnetic signal distribution characteristics at the inner and outer wall welds of the pipeline are calculated. Furthermore, the effects of different excitation intensities on the amplitude growth characteristics of magnetic signals are analyzed and compared. To verify the analysis model, weld detection experiments with different excitation intensities were designed. The results show that both the peak-to-valley values of the normal component and the peak values of the tangential component of the outer wall weld are lower than those of the inner wall weld. Conversely, the peak-to-valley width of the normal component and the peak width of the tangential component are greater than those of the inner wall weld. Additionally, the rate of increase in weak magnetic signal amplitude decreases in a first-order exponential relationship with increasing excitation intensity. The average decay rates of the normal and tangential component amplitude growth rates for the inner wall weld are 34.03% and 27.9%, respectively, while for the outer wall weld, they are 31.75% and 28.01%, respectively. This study contributes to the identification and quantitative assessment of weak magnetic signals in inner and outer wall welds.

Applying (ARFIMA) Model for Forecast the Saudi Stock Market Prices

The interest in the topic of time series forecasting has increased during the recent years and thus appeared specific modern methods, for example Autoregressive Fractional Integrated Moving Average model (ARFIMA), or what is called long memory model, which use fractionally difference (d) instead of integer which used in ARIMA models. In this study we displayed long memory feature and its tests. Our discussion supported by analysing the real time series (daily closing index of the Saudi Arabia Stock Market prices) over period 1/1/2018 to 19/12/2022, including 1240 observations, to make a good model to giving forecast results for future, using statistical tests and statistical software (R- 4.2.2 program). Firstly we checked that the data series was unstable by testing unit root, using Dickey fuller and Philips Perron tests ; and confirmed the presence of a long memory pattern in it by calculating the Hurst exponent (H), then calculated the fractional differential coefficient (d), determined the appropriate models for analysis and prediction, then the best model was chosen among them based on the comparison criteria, then predicted the values for the next 5 days, R - 4.2.2 software was used in all of these tests and predictions Statistical results indicated that the optimal model to represent the data series is ARFIMA (3,0.383,2) which used to predict future values.

Adaptive Evolutionary Computing Ensemble Learning Model for Sentiment Analysis

Standard machine learning and deep learning architectures have been widely used in the field of sentiment analysis, but their performance is unsatisfactory if the input texts are short (e.g., social media posts). Specifically, the accuracy of standard machine learning methods crucially depends on the richness and completeness of the features used to represent the texts, and in the case of short messages, it is often difficult to obtain high-quality features. Conversely, methods based on deep learning can achieve better expressiveness, but these methods are computationally demanding and often suffer from over-fitting. This paper proposes a new adaptive evolutionary computational integrated learning model (AdaECELM) to overcome the problems encountered by traditional machine learning and deep learning models in sentiment analysis for short texts. AdaECELM consists of three phases: feature selection, sub classifier training, and global integration learning. First, a grid search is used for feature extraction and selection of term frequency-inverse document frequency (TF-IDF). Second, cuckoo search (CS) is introduced to optimize the combined hyperparameters in the sub-classifier support vector machine (SVM). Finally, the training set is divided into different feature subsets for sub-classifier training, and then the trained sub-classifiers are integrated and learned using the AdaBoost integrated soft voting method. Extensive experiments were conducted on six real polar sentiment analysis data sets. The results show that the AdaECELM model outperforms the traditional ML comparison methods according to evaluation metrics such as accuracy, precision, recall, and F1-score in all cases, and we report an improvement in accuracy exceeding 4.5%, the second-best competitor.

An Efficiency Evaluation and Driving Effect Analysis of the Green Transformation of the Thermal Power Industrial Chain: Evidence Based on Impacts and Challenges in China

The high carbon emissions and pollution of China’s thermal power industry chain have exacerbated environmental and climate degradation. Therefore, accelerating the green transformation process is of great significance in promoting the sustainable development of enterprises. This study selected 30 listed thermal power enterprises in China as research objects, analyzed their data from 2018 to 2022, set targeted input–output indicators for different stages, and used a two-stage dynamic data envelopment analysis (DEA) model to evaluate and measure the efficiency of the green transformation of Chinese thermal power enterprises. In addition, this study also uses the logarithmic mean Divisia index (LMDI) method to analyze the driving effects of green transformation. The results indicate that in terms of overall efficiency, there is a significant difference in the overall performance of these 30 thermal power enterprises, with a large difference in average efficiency values. Efficiency values are related to enterprise size. In terms of stage efficiency, the average efficiency value of thermal power enterprises in the profit stage was significantly higher than that in the transformation stage, and the profitability of Chinese thermal power enterprises was better. In terms of sub-indicator efficiency, the efficiency of each indicator shows a “U”-shaped trend, and there is a certain correlation between the operating costs and revenue of thermal power enterprises, the market value of green transformation, and related indicators. In addition, the most important factor affecting the efficiency of green transformation is the sewage cost they face, whereas their operational capabilities have the least impact on their green transformation. In this regard, thermal power enterprises should increase their investment in the research and development of key technologies for thermal power transformation and continuously optimize their energy structure. The government will increase financial support for thermal power green transformation enterprises and correspondingly increase emission costs.

Model for the Failure Prediction Mechanism of In-Service Pipelines Based on IoT Technology

With the rapid increase in pipeline mileage in China, the accurate prediction of corrosion issues in in-service pipelines has become crucial for ensuring safe pipeline operation. Traditional pipeline leakage monitoring methods are significantly limited by human factors and equipment precision, making it challenging to predict and identify leakage points accurately. Therefore, aligned with the trend of intelligent pipeline development, this study aims to construct a failure pressure prediction mechanism model for corroded pipelines based on IoT technology. This model leverages intelligent sensing and prediction to assess the safety status of corroded pipeline sections. Ultrasonic phased array technology detects specific corrosion points and detailed defect parameters within pipeline sections. The parameters are then utilized in the Simdroid domestic finite element analysis model to simulate the ultimate burst pressure of the pipeline. A single-variable approach is employed to analyze the sensitivity of different parameters to the pipeline’s ultimate burst pressure, with the minimum burst pressure point of multi-point corroded sections selected as the overall segment failure pressure. Finite element simulation data are integrated into a neural network database to predict the pipeline failure pressure. The real-time operational data of the pipeline are monitored using negative-pressure wave sensing. The operational pressure of the corroded points is compared with the algorithm-predicted failure pressure; if the values approach a critical threshold, an alarm is triggered. Moreover, the remote control terminals evaluate the pipeline’s self-rescue time, providing a buffer for pipeline leakage self-rescue. The failure prediction mechanism model for in-service pipelines was applied to the Fujian–Guangdong branch of the West–East Gas Pipeline III to verify its accuracy and feasibility. The research results offer technical support for the maintenance and emergency repair of pipeline leakage scenarios, leveraging intelligent pipeline technology to reduce costs and increase the efficiency of pipeline operations, thereby supporting the sustainable development of China’s oil and gas pipelines with theoretical and technical backing.

A novel scoring system based on magnetic resonance imaging for the prediction of the difficulty of ultrasound-guided high-intensity focused ultrasound ablation for uterine fibroids

Objective To develop a novel scoring system based on magnetic resonance imaging (MRI) for predicting the difficulty of ultrasound-guided high-intensity focused ultrasound (USgHIFU) ablation for uterine fibroids. Materials and methods A total of 637 patients with uterine fibroids were enrolled. Sonication time, non-perfused volume ratio (NPVR), and ultrasound energy delivered for ablating 1 mm3 of fibroid tissue volume (E/V) were each classified as three levels and assigned scores from 0 to 2, respectively. Treatment difficulty level was then assessed by adding up the scores of sonication time, NPVR and E/V for each patient. The patients with score lower than 3 were categorized into low difficulty group, with score equal to or greater than 3 were categorized into high difficulty group. The potential predictors for treatment difficulty were compared between the two groups. Multifactorial logistic regression analysis model was created by analyzing the variables. The difficulty score system was developed using the beta coefficients of the logistic model. Results Signal intensity on T2WI, fibroid location index, largest diameter of fibroids, abdominal wall thickness, homogeneity of the signal of fibroids, and uterine position were independent influencing factors for the difficulty of USgHIFU for uterine fibroids. A prediction equation was obtained: difficulty score = 17 × uterine position (anteverted =0, retroverted =1)+71 × signal intensity (hypointense = 0, isointense/hyperintense = 1) +8 × enhancement (homogenous = 0, heterogeneous = 1)+25×(largest diameter of fibroids-20) +35 × (fibroid location index −0.2) +1×(abdominal wall thickness −5). Conclusions This scoring system established based on MRI findings can be used to reliably predict the difficulty level of USgHIFU treatment of uterine fibroids.

Exploring the Link between Seafood and Traditional Fish Sauces, and Blood Pressure: A Bandar Kong Cohort Study

Background: High blood pressure (BP) is a risk factor for various health conditions, and dietary factors have been linked to the development of high blood pressure. Scientific studies have shown that consuming marine food, particularly oily fish, can be beneficial in reducing blood pressure levels. This study aims to investigate the relationship between traditional fish sauces and seafood consumption, and blood pressure levels in Bandar Kong Cohort study. Methods: Cross-sectional analyses were performed using data from 2823 volunteers, collected through questionnaires and BP measurements following a standardized protocol. Age-adjusted and multivariate associations between BP and seafood consumption were estimated using multiple linear regressions. Three models of analysis were used: model 1, daily energy intake; model 2, daily energy intake along with age; and model 3, daily energy intake, age, education, body mass index (BMI), residence, occupation, marital status, using hookah, socio-economic status, physical activity, and family history of hypertension. Results: Among 2823 participants (age; 46.88±8.94 yrs.), 1138(40.3%) men and 1685(59.7%) women, 693 (24.5%) individuals had hypertension, (n=288) accounting for 25.3% men and (n=405) 24.0% women. After adjusting for covariates, the study found significant negative relationships between the consumption of salted fish-water, Mahyaveh and Suragh (traditional Iranian fish sauces), and BP. Conclusions: This study suggests that consuming traditional Iranian fish sauces, Mahyaveh and Suragh, is associated with lower BP levels. This study sheds light on potential benefits of traditional dietary choices for regulating blood pressure.

A dual methodology employing ion-pair chromatography and built-in UV spectrophotometry for quantifying recently approved combination of mometasone and indacaterol in a novel combined metered dose inhaler: assessing the greenness, carbon footprint, blueness, and whiteness

Developing analytical techniques that align with green and sustainable chemistry principles is crucial in today's scientific landscape. This work introduces two innovative approaches for the simultaneous quantification of indacaterol (IND) and mometasone (MOM), a recently approved combination therapy for chronic obstructive pulmonary disease. These methods—rapid isocratic ion pair chromatography (IPC) and UV–visible spectrophotometry—demonstrate improved environmental sustainability, cost-effectiveness, and versatility compared to existing techniques. The optimized 4-min IPC method achieved excellent resolution (retention times 2.18 ± 0.1 min for IND and 3.95 ± 0.1 min for MOM), peak symmetry, and sensitivity. It utilizes a low-cost ion pair mobile phase of acetonitrile and acidified water containing 0.025% sodium dodecyl sulfate (50:50% v/v), making it suitable for laboratories with standard chromatographic instruments. The spectrophotometric approach offers two procedures: first derivative and ratio derivative methods. These serve as simplified, low-cost alternatives for resource-limited laboratories without access to advanced instruments. Both techniques feature simplified protocols that minimize extraction and fractionation steps. Comprehensive validation confirmed outstanding accuracy (98–102%) and precision (%2 <). Sustainability assessments using ComplexGAPI, AGREE, carbon footprint, BAGI, and RGB12 tools demonstrated enhanced environmental performance compared to existing methods. The IPC and spectrophotometry methods achieved greenness scores of 0.81 and 0.85, respectively, surpassing the 0.63–0.67 range of reported techniques. Additionally, they showed lower carbon footprints of 0.035 and 0.022 kg CO2 equivalent emissions per sample, compared to 0.079–0.092 kg for conventional procedures. The application of novel "blueness" and "whiteness" concepts using BAGI and RGB12 algorithms further confirmed superior sustainability, with scores of 87.5 & 90 for blueness and 88.1 & 89.8 for whiteness. Successfully applied to quantify IND and MOM in combined capsules, this work provides a model for eco-friendly pharmaceutical analysis that maintains high analytical reliability while improving sustainability metrics.

Estimation of the mean exponential survival time under a sequential censoring scheme

The exponential distribution can provide a simple and appealing survival-time model in reliability analysis and life tests. Due to certain experimental designs, time limitations, budgetary requirements, and other reasons, however, only censored data can be obtained. In this paper, we propose a novel estimation procedure for the mean survival time of an exponential distribution under a sequential censoring scheme, which can be treated as a combination of type I censoring and type II censoring. The procedure makes a trade-off between estimation error and sampling cost, using the minimum number of observations. An extensive set of Monte Carlo simulations is conducted to further validate its remarkable performance. To demonstrate the practical applicability, we then implement this newly proposed procedure to assess the reliability of various Backblaze's hard disk models.

Fine-grained sentiment analysis using multidimensional feature fusion and GCN

ABSTRACT The demand for analysing sentiment information in social media data is increasing. However, current fine-grained sentiment analysis methods fail to consider both global and local semantic features simultaneously, leading to the oversight of grammatical information in sentences and an inability to address the issue of polysemy. To address these challenges, we propose a microblog fine-grained sentiment analysis model based on multidimensional feature fusion and graph convolutional neural networks (GCN). Built upon the ALBERT model, we utilize BiLSTM and capsule network models to extract global and local semantic features, thereby capturing bidirectional semantic dependencies and textual positional semantic information. Finally, we employ multi-head self-attention and GCN to select key features and sentence information, ensuring the integrity of fine-grained features. The experimental results indicate that the model outperforms several other models on fine-grained sentiment analysis datasets SMP2020-EWECT, NLPCC2013, NLPCC2014, and the binary classification dataset weibo_senti_100k, achieving accuracies of 80.64%, 67.19%, 71.37%, and 98.43%, respectively.

Generalized model for eigenfrequency analysis of bolted variable-stiffness flanged-cylindrical shells

A general semi-analytical modeling method for the bolted variable stiffness composite cylindrical shell, incorporating precise modeling of the boundary and connecting flanges, is proposed in this paper. The coupling between shells and flanges is achieved through a spring-based penalty function method. A novel bolted joint model for flange connections is introduced, enabling the simulation of non-uniform pressure distributions at the joint interface. The joint model features a stiffness characterization technique that effectively improves adjustability and computational efficiency. The correctness and accuracy of the proposed modeling method are systematically validated through finite element analysis based on the ANSYS commercial software and experimental testing approaches. The series of parameter influence analyses on the bolted variable stiffness composite cylindrical shell with flanges, conducted using the proposed modeling method, accurately demonstrate the influence of geometric dimensions, fiber layup parameters, and bolt parameters on the free vibration behavior of such structures. Furthermore, these analyses aid in a deeper understanding of the dynamic characteristics of such structures, partially highlighting key design considerations and offering valuable guidance for their dynamic optimization design.

Integrated effects of pavement simulation models and scale differences on the thermal environment of tropical cities: physical and numerical modeling experiments

Simulation methods attempt to explain what happens in full-scale environments. However, as simplification procedures, they also have their limitations and opportunities. One of the applications is to use the output data of a physical model to calibrate numerical simulation, or even to use outputs of numerical simulations to analyze urban scale studies. But it is uncertain the error in the interaction between these models. This study aims to analyze the impact of scale analysis and pavements simulation model modification on ambient and surface temperature of asphalt pavement in a physical model of a tropical city street canyons. Therefore, a scaled outdoor experiment was conducted, and a numerical simulation model, using ENVI-met software, was used to investigate the spatiotemporal variation of air and pavement surface temperature, in urban (1:1) and reduced (1:15) scales. For studies on the surface temperature of pavements, within the temperature range of 12 ºC to 37 ºC, it is recommended to calibrate physical models using as input, data derived from numerical simulation models, yielding a mean absolute percentage error (MAPE) of 4.9%. For estimating data in real-world urban scale, within the air temperature range of 15 ºC to 37 ºC, it is proposed to use output data from simulated models in ENVI-met, that presented a mean absolute error (MAE) of ± 0.59 or physical models (MAE = ± 0.66). These results would be useful for the development of urban surface temperatures parametrizations.