Analysis Model Research Articles

Sentiment Analysis of Visitor Reviews on Baturaden Tourist Attraction Using Machine Learning Methods

This study evaluates the performance of four machine learning models: Support Vector Machine (SVM), Random Forest, K-Nearest Neighbors (KNN), and Naive Bayes in analyzing visitor reviews of the Lokawisata Baturaden tourist attraction. Using 5-fold cross-validation, the study aims to determine which machine learning model best suits sentiment analysis on the Baturaden review data. This study was conducted through several stages, including data preprocessing, feature extraction, and the data training process. Case folding, text cleaning, tokenization, stopword removal, and stemming were performed during the data preprocessing stage. The feature extraction method used was TF-IDF. SMOTE was applied to increase data variation and address the data imbalance in the dataset. The results show that SVM provides the best performance with an accuracy of 0.937, an F1-score of 0.937, a precision of 0.943, and a recall of 0.937. Random Forest also performs well with an accuracy of 0.918 and an F1-score of 0.918, though slightly below SVM. KNN shows the lowest performance with an accuracy of 0.651 and an F1-score of 0.544, while Naive Bayes performs adequately with an accuracy of 0.845 and an F1-score of 0.841. Based on this evaluation, SVM is recommended as the best model for sentiment analysis of reviews, followed by Random Forest as a good alternative. The KNN model is not recommended due to its lower performance, while Naive Bayes can be considered for its speed and simplicity, although its results are not as good as SVM and Random Forest. These conclusions guide the selection of the optimal model to enhance understanding and visitor experience at the Baturaden tourist attraction.

Generation of monthly mean flows using the Gr2m model for trend analysis

Generation of monthly mean flows using the Gr2m model for trend analysis

The effect of age on the relationship between negative life events and current and new depressive episodes: results from the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil).

There is scarce data concerning the relationship between negative life events (NLEs)* and current and new depressive episodes by age. Cross-sectional (baseline) and prospective analyses (4-year/8-year follow-ups) were performed in 15,105 civil servants in 6 cities in Brazil classified according to age strata at baseline: 1st(35-44), 2nd(45-54), 3rd(55-64), and 4th(65-74) years. The independent variable was NLEs in the last year (robbery, hospitalization, death of a relative, financial hardship, and rupture of a love relationship) collected at baseline. The dependent variable was depressive episodes assessed at baseline (current), and 4-year/8-year follow-ups (new). We built logistic (Odds Ratio[OR];95 % Confidence Interval[CI]) in the cross-sectional analysis and Poisson regression models (Relative Risk[RR], [CI]) in the prospective analysis. Robbery, hospitalization, financial hardship, rupture of a relationship were associated with current depressive episodes concentrated in the 35-44 age range, while for new depressive episodes, the association of hospitalization, death of a relative, and financial hardship were concentrated in the 45-54 age stratum. Financial hardship was associated with current depressive episodes in all age-strata: 1st: OR, 2.77(CI, 1.83-4.19); 2nd: OR, 1.71(CI, 1.26-2.34); 3rd: OR, 1.68(CI,1.15-2.46); 4th, OR, 4.07(CI, 1.24-13.34), and new depressive episodes: 1st: RR, 1.45(CI, 1.09-1.93); 2nd RR, 1.46(CI: 1.15-1.84); 3rd: RR, 1.97(CI, 1.38-2.82). Rupture of a relationship was associated only with current depressive episode, while death of a relative only with new depressive episode. We cannot rule out the possibility of recall bias, since NLEs` information was self-reported. The association between NLEs with depressive episodes was concentrated in youngers.

Trends in NICE technology appraisals of non-small cell lung cancer drugs over the last decade.

The aim of this study is to analyse the trends in technology appraisals for non-small cell lung cancer (NSCLC) treatments performed by the National Institute for Health and Care Excellence (NICE) over the last ten years. A systematic search was conducted for single technology appraisals of NSCLC drugs in the online NICE database from 2012 to 2022. Search terms used were 'non small cell lung cancer', and 'NSCLC'. Appraisals that were under development or terminated as well as multiple technology appraisals were considered out of scope. In the 30 included appraisals for targeted therapies and immunotherapies within NSCLC, a total of 53 different comparators were included by NICE for 41 assorted indications or subgroups. Partitioned survival models were most frequently used, often including three health states and time horizons of up to 30 years. Throughout the decade the use of indirect comparisons was high and became more established and complex over time. Of all appraisals, 90% positively recommended the treatment for use in the UK. Technology appraisals became more complex over time due to the emergence of targeted therapies and immunotherapies, leading to multiple different indications, subpopulations and comparators that needed to be included in appraisals. Partitioned Survival Analysis (PartSA) models became the cornerstone within NSCLC, with time horizons up to 30 years and over time methods for indirect treatment comparisons became more established. The majority of the appraisals resulted in a positive recommendation for reimbursement.

Sleep-disordered breathing increases mortality in patients with diabetes.

Sleep-disordered breathing (SDB) and diabetes mellitus (DM) are often concomitant; however, data on the impact of SDB on mortality in the population with diabetes remain scarce. The population from the Sleep Heart Health Study, a multicenter prospective observational study representing 5780 patients with polysomnography and mortality data, including 453 patients with DM, was analyzed to assess the impact of SDB variables and the presence of DM on all-cause, cardiovascular disease (CVD), and non-CVD associated mortality. Survival analysis and proportional hazard regression models were used to calculate the adjusted hazard ratios (aHR) for mortality. Patients with DM and the average SpO2 >91.4% had significantly lower all-cause (aHR 0.52, CI 0.34-0.80) and CVD mortality risk (aHR 0.44, CI 0.22-0.87) as compared with patients with SpO2 below this value. Apnea-hypopnea index >31 (aHR 1.58, CI 1.10-2.28) and oxygen desaturation index >13.3 (aHR 1.58, CI 1.10-2.25) were associated with increased all-cause mortality in participants with DM on treatment. Sleep efficiency and proportion of rapid-eye movement (REM) sleep did not have any impact on mortality in patients with DM and thus differed significantly from individuals without DM, where increased all-cause mortality was observed in those with sleep efficiency <81.4% (aHR 0.77, CI 0.68-0.87) or REM sleep <14.9% (aHR 0.78, CI 0.68-0.89). Patients with diabetes on treatment and moderate to severe sleep-disordered breathing experience increased all-cause mortality. Reduced average oxygen saturation predicted both all-cause and cardiovascular death in the population with diabetes.

Rehospitalisation risk by hypnotics class in older patients with heart failure: a cohort study utilizing administrative claims data in Japan

BackgroundStudies comparing the safety of orexin receptor antagonists and other hypnotic types for older patients with heart failure (HF) remain lacking. This study aimed to compare orexin receptor antagonists (suvorexant)...

Extracellular enzyme activity and stoichiometry reveal P limitation in the wild panda habitat of the Qinling Mountains

AbstractSoil microbial extracellular enzymes play crucial roles in soil carbon and nutrient cycling by catalyzing soil biochemical processes. However, the activity and stoichiometry of enzymes involved in the carbon, nitrogen, and phosphorus cycles in the environment with the highest density of wild giant pandas on the southern slopes of the Qinling Mountains is unknown. We have established eight research areas at an elevation of 1090–2621 m. The β‐1,4‐glucosidase (BG), β‐1,4‐N‐acetylglucosaminidase (NAG), and acid phosphatase (APase) in soil samples were measured. We used redundancy analysis and structural equation model to evaluate the driving factors of metabolic restriction of soil microorganisms along the elevational gradient, and four models were used to cross‐evaluate the nutrient restriction status of soil microorganisms. The results showed that most soil physiochemical properties, soil microbial biomass, and microbial extracellular enzymes exhibited a hump‐shaped trend with increasing elevation. Elevation indirectly affected soil enzyme activity and stoichiometry by C, N, and P status. Microorganisms are limited by C at lower and higher elevations but limited by N at medium elevations. These results could help strengthen the conservation and management of the wild panda's natural habitat.

Geospatial foundation models for image analysis: evaluating and enhancing NASA-IBM Prithvi’s domain adaptability

Research on geospatial foundation models (GFMs) has become a trending topic in geospatial artificial intelligence (AI) research due to their potential for achieving high generalizability and domain adaptability, reducing model training costs for individual researchers. Unlike large language models, such as ChatGPT, constructing visual foundation models for image analysis, particularly in remote sensing, encountered significant challenges such as formulating diverse vision tasks into a general problem framework. This paper evaluates the recently released NASA-IBM GFM Prithvi for its predictive performance on high-level image analysis tasks across multiple benchmark datasets. Prithvi was selected because it is one of the first open-source GFMs trained on time-series of high-resolution remote sensing imagery. A series of experiments were designed to assess Prithvi’s performance as compared to other pre-trained task-specific AI models in geospatial image analysis. New strategies, including band adaptation, multi-scale feature generation, and fine-tuning techniques, are introduced and integrated into an image analysis pipeline to enhance Prithvi’s domain adaptation capability and improve model performance. In-depth analyses reveal Prithvi’s strengths and weaknesses, offering insights for both improving Prithvi and developing future visual foundation models for geospatial tasks.

Exploration of the Indoor Layout Optimization Model in Computer-Aided Visual Analysis

Exploration of the Indoor Layout Optimization Model in Computer-Aided Visual Analysis

A Model for Institutional Capacity Analysis for Education Sector at Regency Level in Indonesia: A Case Study on the Office of Education the Regency of Sleman

A Model for Institutional Capacity Analysis for Education Sector at Regency Level in Indonesia: A Case Study on the Office of Education the Regency of Sleman

SAD: semi-supervised automatic detection of BOLD activations in high temporal resolution fMRI data.

Despite the prevalent use of the general linear model (GLM) in fMRI data analysis, assuming a pre-defined hemodynamic response function (HRF) for all voxels can lead to reduced reliability and may distort the inferences derived from it. To overcome the necessity of presuming a specific model for the hemodynamic response, we introduce a semi-supervised automatic detection (SAD) method. The proposedSADmethod employs a Bi-LSTM neural network to classify high temporal resolution fMRI data. Network training utilized an fMRI dataset with 75-ms temporal resolution in an iterative scheme. Classification performance was evaluated on a second fMRI dataset from the same participant, collected on a different day. Comparative analysis with the standard GLM approach was conducted to evaluate the cooperativeeffectiveness of the SAD method. The SAD method performed well based on the classification scores: true-positive rate = 0.961, area under the receiver operating curve = 0.998, true-negative rate = 0.99, F1-score = 0.979, False-negative rate = 0.038, false-discovery rate = 0.002, false-positive rate = 0.002 at 75-ms temporal resolution. SAD can detect hemodynamic responses at 75-ms temporal resolution without relying on a specific shape of an HRF. Future work could expand the use cases to include more participants and different fMRI paradigms.

Cross-cultural adaptation and psychometric properties of the Chinese version of the German social support questionnaire among older adults with chronic diseases.

Social support is widely recognized as a protective factor against psychological distress, especially for the elderly with chronic diseases. Therefore, effective tools for measuring social support are of great significance for both research and clinical practice. This study aims to cross-culturally adapt the German Social Support Scale (F-SozU) into Chinese and assess its psychometric properties among older adults with chronic diseases. A cross-sectional study was conducted in a Chinese tertiary hospital. Four hundred ninety-six older adults that were approached via a convenience sampling method completed the Chinese version of the F-SozU. Content validity was evaluated using the two-round Delphi method. Psychometric properties, including item analysis, content validity, structure validity, convergent and discriminant validity and reliability were examined. The final version of C-F-SozU is a three-factor structure consisting of 23 items. All indicators of item analysis are acceptable. Adequate content validity was ensured by the expert panel (I-CVI = 0.80-1.00, S-CVI = 0.965) and participants. The confirmatory factor analysis model revealed that the factor structure of the C-F-SozU fitted the original scale (χ 2 /df = 2.088, CFI = 0.998, GFI = 0.943, TLI = 0.997, IFI = 0.998, RMSEA = 0.064 and SRMR = 0.043). The total Cronbach's α was 0.956, and the test-retest reliability coefficient was 0.887. The convergent validity (average variance extracted = 0.517-0.995) and discriminant validity were found to be satisfactory. No floor/ceiling effect was found. The 23-item C-F-SozU demonstrates robust reliability and validity, rendering it a valuable instrument for evaluating social support among older adults with chronic diseases in China. The three-factor structure of the scale allows for a more detailed assessment of the social support, with the scores of each dimension and the total score being of significant reference value. More comprehensive studies may be required to confirm its effectiveness and applicability.

Building a Defence Manpower Planning Tool Using Survival Analysis Models

ABSTRACT A vibrant economy results in strong competition in the marketplace for highly skilled and versatile workforce. Therefore, the retention of such workers is an important issue for most organisations, including the Royal Navy, which is sensitive to economic pressures from the civilian economy. This paper considers early exits from the Royal Navy as the termination of an individual’s career in the naval force at any time before the termination of their contractual agreement with the organisation. The paper uses a survival analysis model to forecast/predict early exits from the Royal Navy and help build a defence manpower planning tool to manage retentions. At the micro-level, it provides two sets of predictions, the first on who is more likely to leave early, and the second on when that event is most likely to occur. At the aggregate level, the predictive model provides a management tool for performing stress testing in relation to potential changes in economic and labour market conditions and how the heterogeneous skilled workforce within the organisation is likely to respond to different changes.

Favipiravir for COVID-19 in adults in the community in PRINCIPLE, an open-label, randomised, controlled, adaptive platform trial of short- and longer-term outcomes.

Evidence for the effect of favipiravir treatment of acute COVID-19 on recovery, hospital admissions and longer-term outcomes in community settings is limited. In this multicentre. open-label, multi-arm, adaptive platform randomised controlled trial participants aged ≥18 years in the community with a positive test for SARS-CoV-2 and symptoms lasting ≤14 days were randomised to: usual care; usual care plus favipiravir tablets (loading dose of 3600mg in divided doses on day one, then 800mg twice a day for four days); or, usual care plus other interventions. Co-primary endpoints were time to first self-reported recovery and hospitalisation/death related to COVID-19, within 28 days, analysed using Bayesian models. Recovery at six months was the primary longer-term outcome. ISRCTN86534580. The primary analysis model included 8811 SARS-CoV-2 positive mostly COVID vaccinated participants, randomised to favipiravir (n=1829), usual care (n=3256), and other treatments (n=3726). Time to self-reported recovery was shorter in the favipiravir group than usual care (estimated hazard ratio 1·23 [95% credible interval 1·14 to 1·33]), a reduction of 2·98 days [1·99 to 3·94] from 16 days in median time to self-reported recovery for favipiravir versus usual care alone. COVID-19 related hospitalisations/deaths were similar (estimated odds ratio 0·99 [0·61 to 1·61]; estimated difference 0% [-0·9% to 0·6%]). 14 serious adverse events occurred in the favipiravir group and 4 in usual care. By six months, the proportion feeling fully recovered was 74·9% for favipiravir versus 71·3% for usual care (RR = 1·05, [1·02 to 1·08]). In this open-label trial in a largely vaccinated population with COVID-19 in the community, favipiravir did not reduce hospital admissions, but shortened time to recovery and had a marginal positive impact on long term outcomes.

Improving energy-related efficiency towards SDG7 in China: What role does energy poverty play?

The improvement of energy-related efficiency has promoted energy-sustainable development in China; however, it may be inhibited by energy poverty. This paper constructs an index system to measure China's energy poverty. Subsequently, we evaluate energy-related efficiency towards SDG7 comprised of energy production efficiency and energy sustainable utilization efficiency using a dynamic two-stage data envelopment analysis model. Furthermore, we adopt fixed effect models to quantify the impact of energy poverty on improving energy-related efficiency and the underlying mechanism. The findings indicate that: (1) China's energy poverty situation gradually improves from 2011 to 2020. The energy-related system's performance is medium due to the poor performance in its energy production stage in 30 provinces during 2011-2020. Energy-related efficiency is greater in areas with a higher response to SDG7. (2) The impact of energy poverty on improving energy-related efficiency is significantly negative. Addressing energy poverty significantly improves energy-related efficiency in regions characterized by a high energy poverty index and low energy-related efficiency. However, it bears no impact in regions with a low energy poverty index and high energy-related efficiency. (3) The moderating effect of technological innovation effectively improves energy-related efficiency, whereas the marketization level has an inverse effect. Consequently, this paper suggests several policy implications.

A Study of an Integrated Analysis Model with Secondary Flow for Assessing the Performance of a Micro Turbojet Engine

The objective of this study is to implement a more realistic integrated analysis model for micro gas turbines by incorporating secondary flow and combustion efficiency into the existing model, which includes main engine components such as the compressor and turbine, and to validate this model by comparing it with test results. The study was based on the JetCat P300-RX, which has a maximum thrust level of 300 N. Simulations were performed using ANSYS CFX, employing the κ-ω SST turbulence model and a mixing plane interface between individual components. The eddy dissipation model (EDM), with a combustion efficiency of 90%, was used as the combustion model. A user subroutine was also applied for the power matching of the compressor and turbine to calculate the fuel flow rate in each iteration. For secondary flow, it was assumed that 3% of the total air flow rate would flow through the secondary path and be applied to the compressor and turbine. Simulations were conducted over a range of 30,000 to 104,000 RPM, with ground conditions evaluated, including altitude-simulated conditions. To validate the analysis model, engine performance metrics such as pressure ratio, air flow rate, fuel flow rate, and exhaust gas temperature (EGT) were compared with test results. The results demonstrated that errors were less than 5% for most engine performance metrics, except for EGT and fuel flow. The discrepancy in EGT was attributed to differences in the sensing methods, while the variation in fuel flow was found to be due to the lubrication system and losses due to the secondary air flow. Consequently, this study confirmed that the integrated simulation model accurately predicts engine performance. The results indicate that the integrated simulation model provides a more realistic prediction of overall engine performance compared to previous studies. Therefore, it can evaluate detailed thermo-fluid properties without the need for component performance maps, enhancing performance evaluation and analysis.

Predicting Financial Fraud through Pressure, Opportunity, Rationalization and Internal Control (Case Study on Food and Beverage Companies Listed on the IDX)

The Empirical Study of Food and Beverage Companies Listed on the Indonesia Stock Exchange in 2017-2019 aims to explain and forecast the impact of pressure, opportunity, rationalization, and internal control on financial statements that are false. Because it stresses verifying theories using numerical measurements of research variables and statistical data analysis, this study takes a quantitative approach. The study's population consisted of 18 companies that were listed in the food and beverage sub-sector between 2013 and 2017 and were part of the Indonesia Stock Exchange (IDX). Multiple linear regression analysis, often known as multiple regression, is the analysis model that is used to examine how the independent variables affect the dependent variable. The analysis's findings demonstrate how pressure, opportunity, rationalization, and internal control all significantly impact financial statement fraud at the same time. Internal control has little impact on fake financial statements, while pressure, opportunity, and rationalization all have a considerable impact.

Towards a Deep Learning-Powered Herbarium Image Analysis Platform

Global digitization efforts have archived millions of specimen scans worldwide in herbarium collections, which are essential for studying plant evolution and biodiversity. ReColNat hosts, at present, over 10 million images. However, analyzing these datasets poses crucial challenges for botanical research. The application of deep learning in biodiversity analyses, particularly in analyzing herbarium scans, has shown promising results across numerous tasks (Ariouat et al. 2023, Ariouat et al. 2024, Groom et al. 2023, Sahraoui et al. 2023). Within the e-Col+project (ANR-21-ESRE-0053), we are developing multiple deep learning models aimed at identifying plant morphological traits. We have developed pipelines and models for cleaning, analyzing, and transforming herbarium images, including models for: i) detecting non-vegetal elements, such as barcodes, envelopes, labels, etc.; ii) detecting plant organs, including leaves, flowers, fruits, etc.; and iii) segmenting to recognize plant parts for image cleaning. We are also developing models for classification tasks related to various morphological traits. To validate these models, improve their generalization, and make them easily usable by end-users, deploying them within a generic platform is crucial. The generic platform called PlantAI, currently under development by the e-Col+ project, should enable easy deployment during development for testing and allow users to load annotations for new traits in order to train a model and add it to the existing catalog. The platform is based on a microservice architecture, allowing users to upload images, create custom datasets, and access various AI models for image analysis. The platform is composed of four main modules, as illustrated in Fig. 1. The first module is the collaborative workspace manager, which allows users to create projects and image datasets and invite other users to collaborate on a project. The second module is the navigation interface and dashboards. This module integrates a search engine using metadata and AI annotations, a navigation interface between projects, datasets, and specimens, as well as dashboards for analysis across datasets, specimens, and AI models. The third module is the dataset manager, which handles metadata and annotations associated with the specimens. These annotations can be produced either by expert users or by AI models. The fourth module is the AI models management module, so that models can be used to generate AI annotations of specimen. During the development lifecycle of an AI model, users can create datasets and annotate them with AI models. These annotations can be in two possible states: validated by experts and non-validated. Users collaborating on a project can indicate errors in the model predictions and leave comments to explain their evaluations. These corrections made by experts can be used to retrain the models and thus improve their performance. This platform, will be highly beneficial for botanists, enhancing the efficiency and effectiveness of biodiversity analyses from herbarium scans. We aim to provide users with a catalog of AI models through this platform and allow them to import their own datasets with their own annotations regarding traits of their choice. Users will be able to select a model from the AI model catalog and train it using their dataset. Ultimately, the model obtained from this training will be automatically deployed to be available for AI annotation. The annotations produced by this model will be automatically available in the filtering and navigation interface, thus allowing for dynamic and automatic integration of the AI annotations into the navigation interface.

Application of Machine Learning Models for Short-term Drought Analysis Based on Streamflow Drought Index

Application of Machine Learning Models for Short-term Drought Analysis Based on Streamflow Drought Index

WD-1D-VGG19-FEA: An Efficient Wood Defect Elastic Modulus Predictive Model

As a mature non-destructive testing technology, near-infrared (NIR) spectroscopy can effectively identify and distinguish the structural characteristics of wood. The Wood Defect One-Dimensional Visual Geometry Group 19-Finite Element Analysis (WD-1D-VGG19-FEA) algorithm is used in this study. 1D-VGG19 classifies the near-infrared spectroscopy data to determine the knot area, fiber deviation area, transition area, and net wood area of the solid wood board surface and generates a two-dimensional image of the board surface through inversion. Then, the nonlinear three-dimensional model of wood with defects was established by using the inverse image, and the finite element analysis was carried out to predict the elastic modulus of wood. In the experiment, 270 points were selected from each of the four regions of the wood, totaling 1080 sets of near-infrared data, and the 1D-VGG19 model was used for classification. The results showed that the identification accuracy of the knot area was 95.1%, the fiber deviation area was 92.7%, the transition area was 90.2%, the net wood area was 100%, and the average accuracy was 94.5%. The error range of the elastic modulus prediction of the three-dimensional model established by the VGG19 classification model in the finite element analysis is between 2% and 10%, the root mean square error (RMSE) is about 598. 2, and the coefficient of determination (R2) is 0. 91. This study shows that the combination of the VGG19 algorithm and finite element analysis can accurately describe the nonlinear defect morphology of wood, thus establishing a more accurate prediction model of wood mechanical properties to maximize the use of wood mechanical properties.