CHAID Research Articles

Determining the Type of Charolais Cows by Chaid Analysis

Factors influencing diagonal body length of Charolais cows (n=311) were analysed by the CHAID (Chi-squared Automatic Interaction Detector) method and are presented in this paper. CHAID algorithm groups the dependent variables (Y) so that the variance within groups is as low as possible and the variance between groups is as high as possible. The hierarchy of explanatory variables in this study was shown by the order of live weight, height at withers and age of the cows. The 1st, 2nd and 3rd nodus were formed by the live weight, according to the following categories: ≤600 kg, 600-645 kg, >645 kg. Diagonal body length increased parallel with the live weight categories. The 158 cows forming the first nodus were divided into two separate groups by the program, based on height at withers: 4th nodus (n=50, ≤ 128 cm); 5th nodus (n=108, >128 cm). Finally, the 5th nodus was separated – according to the age – into two groups in the last phase of the analysis: 6th nodus (n=53, ≤ 5.35 age, 172 cm); 7th nodus (n=55, > 5.35 age, 177 cm). Grouping individuals within the population would help to identify cow types, therefore CHAID analysis may be a possible useful method in beef cattle breeding.

A machine learning-based protocol to support visual tree assessment and risk of failure classification on a university campus

Tree failure risk assessment involves visually evaluating trees by considering three essential factors: identifying potential targets that may be affected if the tree falls, assessing the potential consequences of the fall, and determining the likelihood of tree failure. This assessment was used to evaluate the safety of trees in a study area at Thammasat University Rangsit Center, Thailand. In two priority-selected areas for tree risk management, 3659 trees representing 139 species were assessed, and to understand the spatial patterns of tree health conditions and risks, the study employed a GIS-based mapping methodology to manage tree inventory and analyze the spatial patterns of tree health conditions and risks. A decision tree protocol based on the chi-squared automatic interaction detector (CHAID) algorithm, which employs machine learning, was used to evaluate the risk of tree failure. Our study successfully identified seven variables that are crucial in assessing the risk of tree failure. According to the findings, the overall accuracy rate of failure risk classification was 87.35 %, and of all the trees evaluated, 280 trees (7.65 % of the total) representing 34 different species were at high risk. It is recommended to start the assessment process by evaluating important variables such as tree cavities, pest infestations, mechanical damage, dead branches, and epicormic growth. Machine learning protocols, integrated with GIS, are shown to be effective, spatially-explicit, decision-support tools for detecting tree failure potential and assessing risk ratings. Application of these tools improves tree risk management practices.

Clinical characteristics, risk factor analysis and peripheral blood cell changes for early warning of multidrug-resistant bacteria (MDR) infection in elderly patients.

To explore peripheral blood indicators that may serve as early indicators for multidrug-resistant bacteria (MDR) infections in this demographic, with the goal of providing reference suggestions for the clinical prevention of MDR infections in elderly inpatients. Clinical data of patients were divided into the MDR-infected group (n = 488) and the MDR-uninfected group (n = 233) according to the results of drug sensitivity experiments, risk factors for MDR infection, and peripheral blood indicators related to MDR infections were analyzed using univariate and multivariate logistic regression in conjunction with the construction of a Chi-squared automatic interaction detector (CHAID) decision tree model, considering statistical significance at p < .05. Of 721 patients, 488 multidrug-resistant strains were identified. Among them, with Staphylococcus spp. the most prevalent in 148 strains. The most frequent detection of MDR occurred in puncture fluid samples (167 cases). Univariate and multivariate regression analyses revealed that prolonged hospitalization, use of antibiotics preadmission, duration of antibiotics, invasive procedures or recent surgery, and coexisting lung disease were independent risk factors for contracting MDR. Subsequent analysis comparing the aforementioned influences with peripheral blood cells revealed associations between the number of antibiotic treatment days and increased neutrophil-to-lymphocyte ratio (NLR), platelet count-to-lymphocyte ratio (PLR), neutrophils, decreased lymphocytes, and increased eosinophils; preadmission antibiotic use correlated with increased PLR, NLR, neutrophils, and decreased lymphocytes; and invasive manipulation or surgery correlated with increased PLR and NLR. Elevated NLR, PLR, neutrophils, lowered lymphocytes, and eosinophils may serve as early indicators of MDR infections in elderly hospitalized patients.

Predicting mortality amongst Jordanian men with heart attacks using the chi-square automatic interaction detection model.

Background: One of the most complicated cardiovascular diseases in the world is heart attack. Since men are the most likely to develop cardiac diseases, accurate prediction of these conditions can help save lives in this population. This study proposed the Chi-Squared Automated Interactive Detection (CHAID) model as a prediction algorithm to forecast death versus life among men who might experience heart attacks. Methods: Data were extracted from the electronic health solution system in Jordan using a retrospective, predictive study. Between 2015 and 2021, information on men admitted to public hospitals in Jordan was gathered. Results: The CHAID algorithm had a higher accuracy of 93.72% and an area under the curve of 0.792, making it the best top model created to predict mortality among Jordanian men. It was discovered that among Jordanian men, governorates, age, pulse oximetry, medical diagnosis, pulse pressure, heart rate, systolic blood pressure, and pulse pressure were the most significant predicted risk factors of mortality from heart attack. Conclusion: With heart attack complaints as the primary risk factors that were predicted using machine learning algorithms like the CHAID model, demographic characteristics and hemodynamic readings were presented.

Comparison of Chi-Square Automatic Interaction Detector (CHAID) and Random Forest Methods in the Classification of Household Poverty Status in Central Java

Central Java was in second position as the province with the highest number of poor people in Indonesia in March 2020. Poverty alleviation efforts have been carried out, but many are still not yet on target. The purpose of this study was to model the classification of household poverty status in Central Java using CHAID and random forest methods and compare the two methods. The data used in this study is data from the 2020 National Socioeconomic Survey (SUSENAS) conducted by the Central Bureau of Statistics (BPS) for Central Java. The number of poor households is much less than non-poor households. Therefore, Synthetic Minority Oversampling Technique (SMOTE) was performed to handle unbalanced data. The random forest method produced better classification performance than the CHAID method with accuracy, sensitivity, specificity, and AUC of 93,95%, 98,43%, 89,92%, and 0,9417, respectively. The important variables that build the random forest model are the floor area of the house, the age of the head of the household, cooking fuel, the place for the final disposal of feces, and ownership of the place to defecate.

Positional Differences in the Efficacy of Critical End-of-Game Possessions in EuroLeague Basketball

In close basketball games where the final score is within a 5-point difference, end-of-game possessions are very important. However, previous studies have not examined the efficiency of various play types’ possessions regarding player positions. Therefore, the study aimed to identify play type actions during end-of-game possessions across player positions (guard, forward, and big) that directly influence the possession’s outcome. The analysis evaluated 1009 possessions from 100 EuroLeague games from the 2022-2023 and 2023-2024 regular seasons using Sport Scout observation software (SportScout Group, Thessaloniki, Greece). The variables observed included isolation (ISO), pick and roll ball handler (PnRBH), pick and roll screener (PnRSC), off-ball screens (OBS), transition (TR), cuts (CUT), spot-up shooting (SUP), post-ups (PUP), handoffs (HO), inbound plays (IN), and putbacks (PB). Possession efficiency was evaluated based on points per possession (PPP) and the possession’s outcome (positive or negative). Chi-square (χ2) tests compared categorical variables, while the Kruskal–Wallis test assessed PPP differences across player positions. Additionally, the chi-square automatic interaction detector (CHAID) decision tree model was used to classify data and make meaningful predictions for the possession play types. The findings revealed that player positions significantly influenced possession efficiency, with noticeable variations in possession distribution. Decision tree analysis underscored the impact of possession play types on outcomes across player positions. In conclusion, the study highlights the predominant role of guards in endof- game possessions, relying heavily on isolation plays but achieving higher efficiency with teamwork-oriented strategies. Forwards demonstrate effectiveness in offensive rebounding situations and off-ball movement, while big players exhibit efficiency in proximity to the basket.

Body mass index as the main predictor for length of stay in COVID-19 patients with mild and moderate symptoms: a cross-sectional study in COVID-19 emergency hospital in Indonesia

Introduction: COVID-19 emerged as a novel global health crisis. While it has since been downgraded from its status as a public health emergency of international concern, the virus persists as a global health challenge. This study was aimed to analyze the predictors of LoS in COVID-19 patients based on nutritional status and patient characteristics data. Methods: Participants of this study were the COVID-19 patients with mild to moderate symptoms who registered in COVID-19 Emergency Hospital, East Java, Indonesia. This study involved 2850 patients extracted from medical records for further analysis. Body mass index (BMI) was generated following the standard formula measured on the first day of hospitalization. LoS was determined by the number of days of hospitalization. The Chi-square automatic interaction detection (CHAID) algorithm was employed for model building. Results: Most participants were overweight (34.6%) and obese (17.2%). Only a few participants were hospitalized with comorbidities such as hypertension (11.6%) and diabetes mellitus (4.1%). The predictive model of LoS indicated that BMI was the main predictor of COVID-19 LoS, with higher BMI showed to prolong the LoS of mild to moderate symptoms patients. Other than BMI, gender and symptoms were also indicated as COVID-19 LoS predictors. Conclusions: Nutritional status is one of the predictors of LoS in COVID-19 patients. Having higher BMI tends to prolong the LoS, especially in male and having fever. LOS was also seen among those with lower BMI less than 18.5, in patients who had cold and flu.

Use of data mining algorithms to predict body weight of Dorper sheep Breed in two selected villages of Lepelle-Nkumpi local municipality in Limpopo province, South Africa

Sheep farmers in village areas find it difficult to estimate body weight without the weighing tools. The best fitted model for estimation of sheep body weight is not conclusive. Classification and regression tree (CART), Chi-square automatic interaction detection (CHAID), Exhaustive chi-square automatic interaction detection (Ex-CHAID) and Multivariate adaptive regression spline (MARS) were used to predict body weight (BW) of 306 Dorper sheep (female = 244 and male = 62) aged 1 to 2 years from linear body measurement traits such as rump height (RH), rump length (RL), heart girth (HG), withers height (WH), and body length (BL). Goodness of fit criteria's such as Pearson's correlation coefficients (r), coefficient of determination (R2), Akaike's information criterion AIC and adjusted R2 (Adj.R2) were computed for data mining algorithms performance assessment. The results indicated that MARS was the best data mining algorithm with r of 0.979, R2 of 0.958, Adj.R2 of 0.957, RMSE of 2.503 and AIC of 422.25. The MARS model provided the most appropriate predictive capability in the prediction of body weight for Dorper sheep with HG, WH, RH, BL, and sex (male) as the important traits. As a result, MARS data mining algorithm might be used to identify factors that might be vital to improve body weight of Dorper sheep for breeding programs.

Chi-square automatic interaction detection (CHAID) analysis of the use of safety goggles and face masks as personal protective equipment (PPE) to protect against occupational biohazards

BackgroundThis study represents the first Spanish investigation to rigorously evaluate compliance with the use of safety goggles and face masks as essential personal protective equipment (PPE) in companies with workplaces involving exposure to biological agents. ObjectivesThis study aimed to examine the degree of use of face masks and safety goggles as personal protective equipment (PPE), the factors that influence their use, and the profile of workers exposed to occupational biological agents in Spanish companies in the health sector, farming sector, meat industry, waste treatment plants, food industry, and veterinary centers. MethodsWe conducted a cross-sectional descriptive study involving 590 Spanish workers from 51 companies. We developed a 34-item questionnaire to assess workers’ perception of risk related to exposure to biological agents in their workplaces. Among the questions, three were designed to measure the degree of use of key protective equipment in sectors with biological agent exposure: protective gloves, safety goggles or face masks. We only analyzed safety goggles and face masks. We performed various statistical analyses, including Cronbach’s alpha, frequency of endorsement, content validity ratio using Lawshe’s method, varimax rotation, the Kaiser-Meyer-Olkin test, and Bartlett’s sphericity test, to assess the internal consistency and reliability of the questionnaire. Additionally, we employed a chi-square automatic interaction detection (CHAID) segmentation analysis, using workers’ responses regarding their attitudes toward safety goggles and face mask usage as PPE for protection against biological risks, with demographic variables as independent factors. ResultsIn the current study, CHAID analysis revealed that workers exposed to group 2 biological agents used more safety goggles and face shields compared with workers exposed to other groups of biological agents. Moreover, workers in laboratories and the food industry used face masks more than workers of other sectors. ConclusionThe CHAID analysis in the current study indicated that workers exposed to biological agents from both group 2 and group 3 demonstrated satisfactory levels of compliance and utilization of protective masks, surpassing their counterparts in terms of usage. Workers in the food and laboratory industries had subpar compliance with preventive measures, and employees from companies with internal health and safety departments exhibited significant adherence to workplace mask usage, safeguarding themselves against biological risks.

Factors Classifying School Adjustment of Children With and Without Special Needs in Inclusive Preschools

Abstract This study investigated the factors predicting the school adjustment of Turkish preschool children with and without special needs by assessing the school adjustment, social skills, developmental functions, problem behaviors, and cognitive skills of 122 children from 53 preschool classes in 13 public preschools, as well as teacher-child and teacher-parent relationships. Multiple linear regression was used to analyze data to identify the main predictors of school adjustment. The children’s classification structure was then modeled using a Chi-squared automatic interaction detector (CHAID) analysis, considering both their important predictors and school adjustment levels. The findings revealed that the children’s social skills were the most critical predictor of school adjustment. The decision tree created based on the results of the CHAID analysis classified the levels of school adjustment of children with different levels of social skills in terms of parent-teacher relations, developmental functions, and problem behaviors.

Prediction of mortality in young adults with cardiovascular disease using artificial intelligence

&lt;b&gt;Background:&lt;/b&gt; Young mortality is prevalent among patients with cardiovascular disease (CVD). To develop prediction models for CVD mortality in young adults, it is crucial to assess CVD risks. Early detection of cardiac disorders using machine learning algorithms, a branch of artificial intelligence (AI) is crucial for preventing more damage to coronary arteries and saving lives.&lt;br /&gt; &lt;b&gt;Aims:&lt;/b&gt; To predict mortality versus a life outcome among young adults (18-45 years) with CVD using AI technique known as Chi-squared automatic interaction detector (CHAID) algorithms.&lt;br /&gt; &lt;b&gt;Methods:&lt;/b&gt; A large-scale dataset was extracted from the electronic health records of 809 young adult patients diagnosed with CVD using a retrospective design. Information was retrieved regarding young adults from Jordan who were admitted to public health institutions between 2015 and the end of 2021.&lt;br /&gt; &lt;b&gt;Results:&lt;/b&gt; CHAID algorithms were chosen among seven prediction models based on accuracy and area under curve to predict mortality vs life in young individuals (18-45 years old) with CVD. The mortality prediction algorithms started with pulse pressure, then diastolic blood pressure, then ischemic heart disease, and last geographical location.&lt;br /&gt; &lt;b&gt;Conclusions:&lt;/b&gt; CHAID model used in our study indicated how the death rate was classified and distributed among a variety of parameters. As a result, we may argue that AI model could provide additional information on how many aspects are articulated in connection to CVD patient fatality situations.

Pedestrian mobility and University campus accessibility: an analysis of student preferences

Prioritizing pedestrian mobility is crucial to foster urban liveability and sustainability, particularly on university campuses that are experienced by students. This paper examines the University of Catania (Italy) campus, a significant open space facing accessibility challenges. Two surveys were conducted within a Living Lab framework to assess student opinions on campus walking experiences and identify key criteria for promoting walking to and within the campus. The Chi-square Automatic Interaction Detector (CHAID) algorithm and the Analytic Hierarchy Process (AHP) were chosen for the analysis, deriving insights for policy-makers and mobility managers to enhance accessibility and encourage active mobility. The specific case study demonstrates the importance of creating pedestrian-friendly environments to foster sustainable lifestyles by addressing accessibility barriers in large open spaces. However, this research provides a wider framework which is applicable to urban planning in similar contexts, emphasizing the pivotal role of pedestrian-centric design in creating vibrant and sustainable university environments.

“You never farm alone”: Farmer land-use decisions influenced by social relations

ContextFarmer land-use decisions are commonly explained and predicted by socio-demographic, economic, ecological, and psychological factors. However, these models explain only part of the empirical data without accounting for the effect of social relations. ObjectiveThis study aimed at exploring the effect of social relations on farmer land-use decisions using the status-power theory of relations. We hypothesised that farmer land-use decisions are driven by the need to comply with requirements of their salient reference groups, such as family, government, and spiritual beings. MethodsWe undertook a case study in the Mt. Kenya region where we conducted individual interviews and focus group discussions among smallholder farmers. We then used chi-square automatic interaction detection (CHAID) to explore influence of social relations on land-use practices. Results and conclusionsReported social relations were diverse and accounted for land-use decisions that could not be explained by socio-demographic, economic or ecological factors. The results showed that a farmer seemed more likely to choose a land-use option if he/she believed his/her salient reference groups would be pleased with the option. Reconciling social relations with other factors such as farmer's socio-demographic factors and geographic location also had a significant effect on the results. SignificanceInsights into the impact of social relations in farmer land-use decisions can explain the often-heterogeneous decisions and can complement the economic analysis that is the conventional focus in analysis of farmer decisions. An understanding of the effects of social relations can strengthen development of policies that motivate implementation of more sustainable agriculture options.

Physical fitness test performance in firefighter trainees: Differences between graduated and released trainees and predicting academy graduation.

Fitness tests have been previously used to predict academy graduation and highlight specific capacities to be targeted in applicants/trainees to optimise their potential for academy success. To compare the fitness of graduated and released (did not complete academy requirements) firefighter trainees and explore using decision tree analysis to predict academy graduation via fitness tests. Retrospective analysis was conducted on 686 trainees who completed an occupational physical ability test (OPAT): Illinois agility test; push-ups; pull-ups; leg tucks; estimated maximal aerobic capacity (VO2max); backwards overhead 4.54-kg medicine ball throw; 10-repetition maximum deadlift; and 91.44-m farmer's carry. Data were recorded in raw and scaled scores (tests scored from 0-100; maximum OPAT score was 800). Trainees were split into 'graduated' (GRAD; n = 576) or 'released' (REL; n = 110) groups. Mann-Whitney U-tests compared between-group OPAT scores. A decision tree analysis using Chi-square automatic interaction detection was conducted, with raw and scaled scores entered into the analysis. A separate analysis was conducted with only the raw scores. GRAD trainees outperformed REL trainees in all OPAT events (p < 0.001). OPAT total score was the best predictor of academy graduation (p < 0.001), followed by the deadlift score (p = 0.003). Estimated VO2max was the only significant raw score predictor (p < 0.001). GRAD trainees were fitter than the REL trainees. Fitness could predict trainees who graduated from the academy. Overall fitness (OPAT total score), muscular strength (deadlift) and aerobic capacity were important graduation predictors. Training staff could develop these fitness qualities in their trainees to potentially improve fire academy graduation rates.

Comparison of Predictive Performance of Data Mining Algorithms in Predicting Tomato Yield with the A Case Study in Igdir

Among the vegetable species in the world, the plant with the most cultivation area is tomato. Increasing tomato yield is important in terms of contributing more to the world economy, producer’s income and human health. With the advancement in software technologies, the importance of data mining algorithms is increasing due to the fact that these algorithms can produce more sophisticated solutions for regression and classification problems. Determining the factors affecting tomato yield and comparing different data mining algorithms on prediction of tomato yield are the purpose of this study. For this purpose, survey study was conducted with the 105 farmers, selected by Simple Random Sampling Method in Igdir province in 2016. Different data mining algorithms including Classification and Regression Tree, Exhaustive CHAID, Chi-Square Automatic Interaction Detector, Artificial Neural Network Algorithm, Multivariate Adaptive Regression Splines and General Linear Model were developed and compared their predictive performance. MARS decision tree has been able to build a model with greatest predictive accuracy, and the others are respectively ANN, GLM, CART, CHAID and Exhaustive CHAID. In the MARS model, number of irrigation , amount of chemical fertilizer , age of farmer , number of seedlings , education level , soil analysis status , sowing region were found statistically significant (P˂0.05). Preferring the MARS model could give an opportunity to detect factors affecting tomato yield and their interactions with higher accuracy. Moreover, results can be easily interpreted and the rules are understandable.

Medical Conditions in Former Professional American-Style Football Players Are Associated With Self-Reported Clinical Features of Traumatic Encephalopathy Syndrome.

Consensus criteria for traumatic encephalopathy syndrome (TES) specify that at least one core clinical feature of cognitive impairment (CI; e.g., difficulties with memory, executive function) or neurobehavioral dysregulation (ND; e.g., explosiveness, rage, and mood lability) be present and not fully accounted for by other health disorders. Associations between self-reported symptoms that mirror the core clinical features of TES-and how they may be related to concomitant medical conditions-remain unclear. The purpose of this study was to evaluate the association of medical conditions and football exposures with TES clinical features (CI+/-, ND+/-) in 1741 former professional American-style football (ASF) players (age, 57.7 ± 13.9 years; professional seasons, 6.6 ± 3.9 years). Demographics (age, race/ethnicity, current body mass index, age of first football exposure, use of performance-enhancing drugs, position played, and past concussion symptoms), self-reported medical conditions (anxiety, depression, attention-deficit hyperactivity disorder [ADHD], sleep apnea, headache, stroke, hypertension, heart disease, high cholesterol, erectile dysfunction, and low testosterone) were collected. Of 1741 participants, 7.4% were CI+ and/or ND+ (n = 129). Participants who were CI+ or ND+ were more likely to report one or more coexisting medical conditions than participants who did not report CI or ND (odds ratio [OR] = 2.04; 95% confidence interval: 1.25-3.47; p = 0.003). Separate general linear models for each medical condition that adjusted for demographics and football-related factors identified significant associations between ADHD, diabetes, erectile dysfunction, headaches, sleep apnea, anxiety, and low testosterone and CI+ and/or ND+ (ORs = 1.8-6.0). Chi-square automatic interaction detection (CHAID) multi-variable decision tree models that incorporated medical conditions and football exposures accurately differentiated former players meeting either CI or ND clinical criteria from those meeting none (accuracy = 91.2-96.6%). CHAID identified combinations of depression, headache, sleep apnea, ADHD, and upper quartiles of concussion symptom history as most predictive of CI+ and/or ND+ status. CI+ and/or ND+ players were more likely to report medical conditions known to cause cognitive symptoms. Concussion exposure and medical conditions significantly increased the likelihood that a former ASF player would demonstrate cognitive or neurobehavioral dysfunction. Clinicians engaged with this population should consider whether treatable coexisting condition(s) could account for some portion of the clinical picture associated with TES presentation.

Determinants of Physical Activity among Older Adults Through Social Ecological Theory: 2021 Community Health Survey

PURPOSE This study aimed to analyze factors associated with physical activity (PA) in older adults based on social ecological theory.METHODS Secondary analysis was conducted using raw data from the 2021 Community Health Survey. after excluding 129 non-responses in the PA domain, a total of 74,363 individuals were included in the final analysis. A total of eight factors, including personal (level of depression, history of falls), relational (relationships with neighbors, living alone), community (satisfaction with the local community system, safety), and environmental factors (living environment, natural environment), were selected in accordance with the key points of social ecological theory. To analyze social ecological factors related to the PA of older adults, we conducted a decision tree analysis using Chi-square automatic interaction detection (CHAID).RESULTS The average PA level among older adults was approximately 136 minutes, but the mode and median were both 0 minutes. A total of 20.2% of older adults met the physical activity recommendations, while 79.8% did not. According to the first split of the decision tree, living alone was the most relevant factor associated with the PA of older adults, followed by depression and falls. Older adults living alone, with a depression score of 10 or higher, and who had experienced a fall within the past year, were least likely to meet the PA recommendations.CONCLUSIONS This study urgently suggests that PA programs should target older adults living alone, experiencing depression, and falls.

Evidence from the large VALIGA cohort validates the subclassification of focal segmental glomerulosclerosis in IgA nephropathy

Evidence from the Oxford IgA nephropathy (IgAN) cohort supports the clinical value of subclassifying focal segmental glomerulosclerosis lesions (S1). Using the larger Validation in IgA (VALIGA) study cohort, we investigated the association between podocytopathic changes and higher proteinuria, kidney outcome and response to immunosuppressive therapy. All biopsies were evaluated for glomeruli with segmental capillary occlusion by matrix ("not otherwise specified", NOS lesion), simple capsular adhesion without capillary occlusion (Adh), tip lesions, and podocyte hypertrophy (PH). S1 required a NOS lesion and/or Adh. A Chi-Squared Automatic Interaction Detection method was used to identify subgroups of FSGS lesions associated with distinctive proteinuria at biopsy. We assessed survival from a combined event (kidney failure or 50% decline in estimated glomerular filtration rate). Finally, we evaluated within each subgroup if immunosuppression was associated with a favorable outcome using propensity analysis. In 1147 patients, S1 was found in 70% of biopsies. Subclassification found NOS lesions in 44%, Adh in 59%, PH in 13%, and tip lesions in 3%, with much overlap. Four subgroups were identified with progressively higher proteinuria: from lowest, S1 without NOS, S1 with NOS but without Adh/PH, to highest, S1 with NOS and Adh but without PH, and S1 with NOS and PH. These four subgroups showed progressively worse kidney survival. Immunosuppression was associated with a better outcome only in the two highest proteinuria subgroups. Propensity analysis in these two groups, adjusted for clinical and pathological findings, found a significantly reduced time-dependent hazard of combined outcome with corticosteroids. Podocyte hypertrophy and glomeruli with simple adhesions appeared to reflect active lesions associated with a response to corticosteroids, while other S1 lesions defined chronicity. Thus, our findings support subclassifying S1 lesions in IgAN.

Determination of Factors Affecting International Students’ Satisfaction Levels Using CHAID Analysis

Internationalization has recently been a core element for higher education institutions (HEIs) worldwide. Various components provide internationalization in HEIs. However, international students stand out from the others due to providing numerous benefits to host countries. Therefore, HEIs around the world seek to increase the number of international students and their satisfaction levels. This study, inspired from the importance of international students, seeks to determine the factors affecting the satisfaction levels of international students who study at a Turkish state university. To this end, an on-line survey consisting of 25 questions were administered to a total of 330 international students studying at Middle East Technical University (METU). Survey data were evaluated using the CHAID (Chi-Squared Automatic Interaction Detection) analysis. The findings of the analysis indicated that the international students’ implications, study degree and gender played a major role in their satisfaction levels. Considering these influential factors determined by the CHAID analysis will contribute to the internationalization strategies which is developed to attract more international students to HEIs.

Comparative Analysis of Methods for Predicting Brine Temperature in Vertical Ground Heat Exchanger—A Case Study

This research was carried out to compare selected forecasting methods, such as the following: Artificial Neural Networks (ANNs), Classification and Regression Trees (CARTs), Chi-squared Automatic Interaction Detector (CHAID), Fuzzy Logic Toolbox (FUZZY), Multivariant Adaptive Regression Splines (MARSs), Regression Trees (RTs), Rough Set Theory (RST), and Support Regression Trees (SRTs), in the context of determining the temperature of brine from vertical ground heat exchangers used by a heat pump heating system. The subject of the analysis was a public building located in Poland, in a temperate continental climate zone. The results of this study indicate that the models based on Rough Set Theory (RST) and Artificial Neural Networks (ANNs) achieved the highest accuracy in predicting brine temperature, with the choice of the preferred method depending on the input variables used for modeling. Using three independent variables (mean outdoor air temperature, month of the heating season, mean solar irradiance), Rough Set Theory (RST) was one of the best models, for which the evaluation rates were as follows: CV RMSE 21.6%, MAE 0.3 °C, MAPE 14.3%, MBE 3.1%, and R2 0.96. By including an additional variable (brine flow rate), Artificial Neural Networks (ANNs) achieved the most accurate predictions. They had the following evaluation rates: CV RMSE 4.6%, MAE 0.05 °C, MAPE 1.7%, MBE 0.4%, and R2 0.99.