Classification And Regression Tree Technique Research Articles

What we know about antibiotics prescribed by dentists in a Brazilian southeastern state.

The objective of this study was to analyze possible associations between antibiotic dental prescriptions in the public health service, health service characteristics, and social characteristics of the municipalities. Using the register of dispensing in the public health service of a state in the Southeast region of Brazil, in 2017 we analyzed patterns of antibiotic prescriptions by dentists. Data were obtained from the Integrated Pharmaceutical Services Management System (SIGAF, in Portuguese). The outcome variable was the number of Daily Defined Doses (DDD) per 1,000 residents/year in each municipality. CART (Classification and Regression Tree) technique was used to determine the influence of the following variables: populational social characteristics (Human Development Index, Gini Index, the proportion of rural population and proportion of families benefiting from the Bolsa Família Program) and oral health services characteristics (access to individual dental care, number of dentists and oral health teams per 1,000 residents, and percentage of preventive and restorative individual dental procedures. Data analysis was performed using IBM SPSS Statistics 21.0. Antibiotics were the medications most prescribed by dentists in the public health service, with penicillin being the most frequently prescribed class. The average DDD/1,000 residents/year for the 421 municipalities surveyed was 96.54 (range 0.008 and 619.660). Select factors were associated with antibiotic prescriptions including access to individual dental care (Adjusted p-value ≤0.001), a number of oral health teams/1,000 inhabitants (Adjusted p-value=0.001), and Gini Index (Adjusted p-value = 0.046). Access to oral health services and inequality were associated with the use of antibiotics.

Fuzzy Heuristics and Decision Tree for Classification of Statistical Feature-Based Control Chart Patterns

Monitoring manufacturing process variation remains challenging, especially within a rapid and automated manufacturing environment. Problematic and unstable processes may produce distinct time series patterns that could be associated with assignable causes for diagnosis purpose. Various machine learning classification techniques such as artificial neural network (ANN), classification and regression tree (CART), and fuzzy inference system have been proposed to enhance the capability of traditional Shewhart control chart for process monitoring and diagnosis. ANN classifiers are often opaque to the user with limited interpretability on the classification procedures. However, fuzzy inference system and CART are more transparent, and the internal steps are more comprehensible to users. There have been limited works comparing these two techniques in the control chart pattern recognition (CCPR) domain. As such, the aim of this paper is to demonstrate the development of fuzzy heuristics and CART technique for CCPR and compare their classification performance. The results show the heuristics Mamdani fuzzy classifier performed well in classification accuracy (95.76%) but slightly lower compared to CART classifier (98.58%). This study opens opportunities for deeper investigation and provides a useful revisit to promote more studies into explainable artificial intelligence (XAI).

Classification and regression trees for predicting the risk of a negative test result for tuberculosis infection in Brazilian healthcare workers: a cross-sectional study.

Healthcare workers (HCWs) have a high risk of acquiring tuberculosis infection (TBI). However, annual testing is resource-consuming. We aimed to develop a predictive model to identify HCWs best targeted for TBI screening. We conducted a secondary analysis of previously published results of 708 HCWs working in primary care services in five Brazilian State capitals who underwent two TBI tests: tuberculin skin test and Quantiferon®-TB Gold in-tube. We used a classification and regression tree (CART) model to predict HCWs with negative results for both tests. The performance of the model was evaluated using the receiver operating characteristics (ROC) curve and the area under the curve (AUC), cross-validated using the same dataset. Among the 708 HCWs, 247 (34.9%) had negative results for both tests. CART identified that physician or a community health agent were twice more likely to be uninfected (probability = 0.60) than registered or aid nurse (probability = 0.28) when working less than 5.5 years in the primary care setting. In cross validation, the predictive accuracy was 68% [95% confidence interval (95%CI): 65 - 71], AUC was 62% (95%CI 58 - 66), specificity was 78% (95%CI 74 - 81), and sensitivity was 44% (95%CI 38 - 50). Despite the low predictive power of this model, CART allowed to identify subgroups with higher probability of having both tests negative. The inclusion of new information related to TBI risk may contribute to the construction of a model with greater predictive power using the same CART technique.

Türkmen Dağı’ndaki bazı odun dışı orman ürünlerinin potansiyel dağılımı

The aim of this study was to model the potential distribution of non-wood forest products in the Turkmen Mountain of Turkey. Cistus laurifolius, Crataegus monogyna subsp. monogyna, Crataegus orientalis, Rosa canina, Salvia tomentosa and Thymus longicaulis species were selected due to their economic value. Field data were collected from 693 sample plots. Elevation, slope, aspect, parent material, topographic position index, radiation index and bioclimatic variables were used as explanatory variables. Generalized Additive Model (GAM) and Classification and Regression Tree technique (CART) were applied in the modeling processes. The most accurate distribution models for all species were obtained with GAM. Elevation became the most descriptive variable within the distribution models. In addition, parent material, precipitation seasonality (Bio15) and topographical position index played important roles in the model construction process. Visualizations of the distribution models were performed by Geographical Information System software.

Use of a big data analysis technique for extracting HRA data from event investigation reports based on the Safety-II concept

The safe operation of complex socio-technical systems including NPPs (Nuclear Power Plants) is a determinant for ensuring their sustainability. From this concern, it should be emphasized that a large portion of safety significant events were directly and/or indirectly caused by human errors. This means that the role of an HRA (Human Reliability Analysis) is critical because one of its applications is to systematically distinguish error-prone tasks triggering safety significant events. To this end, it is very important for HRA practitioners to access diverse HRA data which are helpful for understanding how and why human errors have occurred. In this study, a novel approach is suggested based on the Safety-II concept, which allows us to collect HRA data by considering failure and success cases in parallel. In addition, since huge amount of information can be gathered if the failure and success cases are simultaneously involved, a big data analysis technique called the CART (Classification And Regression Tree) is applied to deal with this problem. As a result, it seems that the novel approach proposed by combining the Safety-II concept with the CART technique is useful because HRA practitioners are able to get HRA data with respect to diverse task contexts.

Predicting return visits to the emergency department for pediatric patients: Applying supervised learning techniques to the Taiwan National Health Insurance Research Database

Background and objectiveReturn visits (RVs) to the emergency department (ED) consume medical resources and may represent a patient safety issue. The occurrence of unexpected RVs is considered a performance indicator for ED care quality. Because children are susceptible to medical errors and utilize considerable ED resources, knowing the factors that affect RVs in pediatric patients helps improve the quality of pediatric emergency care. MethodsWe collected data on visits made by patients aged ≤18years to EDs from the National Health Insurance Research Database. The outcome of interest was a RV within 3days of the initial visit. Potential factors were categorized into demographics, medical history, features of ED visits, physician characteristics, hospital characteristics, and treatment-seeking behavior. A multivariate logistic regression was used to identify independent predictors of RVs. We compared the performance of various data mining techniques, including Naïve Bayes, classification and regression tree (CART), random forest, and logistic regression, in predicting RVs. Finally, we developed a decision tree to stratify the risk of RVs. ResultsOf 125,940 visits, 6,282 (5.0%) were followed by a RV within 3 days. Predictors of RVs included younger age, higher acuity, intravenous fluid, more examination types, complete blood count, consultation, lower hospital level, hospitalization within one week before the initial visit, frequent ED visits in the past one year, and visits made in Spring or on Saturdays. Patients with allergic diseases and those underwent ultrasound examination were less likely to return. Decision tree models performed better in predicting RVs in terms of area under curve. The decision tree constructed using the CART technique showed that the number of ED visits in the past one year, diagnosis category, testing of complete blood count, and age were important discriminators of risk of RVs. ConclusionsWe identified several factors which are associated with RVs to the ED in pediatric patients. The knowledge of these factors may help assess risk of RVs in the ED and guide physicians to reevaluate and provide interventions to children belonging to the high risk groups before ED discharge.

Forecasting blast-induced ground vibration developing a CART model

Drilling and blasting is an extensively used method for the rock fragmentation in surface mines and tunneling projects. Ground vibration is one of the most important environmental effects produced by blasting operations. In this research work, classification and regression tree (CART), multiple regression (MR), and different empirical models are used to develop predictions for ground vibrations induced by blasting operations conducted in the Miduk copper mine, Iran. To achieve this aim, a number of 86 blasting events were monitored, and the values of peak particle velocity (PPV) in terms of millimeter per second and two effective parameters on the PPV, namely, distance between blast-face and monitoring station in terms of meter and maximum charge used per delay in terms of kilogram, were measured. Performance of models established was evaluated using coefficient of correlation (R2), Nash and Sutcliffe (NS), and root mean square error (RMSE). The results revealed that the CART technique with R2 = 0.95, NS = 0.17, and RMSE = 0.17 provides a better performance when compared with empirical and MR models and has the capacity to generalize.

Classification and regression tree technique in estimating peak particle velocity caused by blasting

Blasting is a widely used technique for rock fragmentation in surface mines and tunneling projects. The ground vibrations produced by blasting operations are the main concern for the industries undertaking blasting operations, which can damage the surrounding structures, adjacent rock masses, roads and slopes in the vicinity. Therefore, proper prediction of blast-induced ground vibrations is essential to demarcate the safety area of blasting. In this research, classification and regression tree (CART) as a rule-based method was used to predict the peak particle velocity through a database comprising of 51 datasets with results of maximum charge per delay and distance from the blast face were fixed as model inputs. For comparison, the empirical and multiple regression (MR) models were also applied and proposed for peak particle velocity prediction. Performance of the proposed models were compared and evaluated using three statistical criteria, namely coefficient of correlation (R2), root mean square error (RMSE) and variance account for (VAF). Comparison of the obtained results demonstrated that the CART technique is more reliable for predicting the peak particle velocity than the MR and empirical models and it can be introduced as a new technique in this field.

Chronic subdural hematoma: Surgical management and outcome in 986 cases: A classification and regression tree approach.

Background:Chronic subdural hematoma (CSDH) is one of the most common clinical entities in daily neurosurgical practice which carries a most favorable prognosis. However, because of the advanced age and medical problems of patients, surgical therapy is frequently associated with various complications. This study evaluated the clinical features, radiological findings, and neurological outcome in a large series of patients with CSDH.Methods:A classification and regression tree (CART) technique was employed in the analysis of data from 986 patients who were operated at Asclepeion General Hospital of Athens from January 1986 to December 2011. Burr holes evacuation with closed system drainage has been the operative technique of first choice at our institution for 29 consecutive years. A total of 27 prognostic factors were examined to predict the outcome at 3-month postoperatively.Results:Our results indicated that neurological status on admission was the best predictor of outcome. With regard to the other data, age, brain atrophy, thickness and density of hematoma, subdural accumulation of air, and antiplatelet and anticoagulant therapy were found to correlate significantly with prognosis. The overall cross-validated predictive accuracy of CART model was 85.34%, with a cross-validated relative error of 0.326.Conclusions:Methodologically, CART technique is quite different from the more commonly used methods, with the primary benefit of illustrating the important prognostic variables as related to outcome. Since, the ideal therapy for the treatment of CSDH is still under debate, this technique may prove useful in developing new therapeutic strategies and approaches for patients with CSDH.

Predicting the Potential Distribution of Plant Diversity in the Yukarıgökdere Forest District of the Mediterranean Region

ABSTRACT: Biodiversity is of crucial importance for ecosystem health and functioning. It is thus one of the most important topics in forest ecology. This study focuses on modelling the spatial distribution of vascular plant diversity (H), richness (S) and evenness (E) in the Yukarigokdere forest district (14667 ha) located in transition zone between Mediterranean and continental climates in the Mediterranean region, Turkey. Species abundance was sampled at 95 sample plots extending from 1000 to 1900 m a.s.l. in order to calculate the response variables (H, S, E). Climatic and topographical data served as explanatory variables. Multiple regression (ML) and classification and regression tree technique (CART) were used to obtain the distribution models of response variables. The better results were provided by CART. That is why visual assessments of the predicted values of the response variables based on (CART) were performed. The explained variances of the model trees were 63.4, 63.1 and 44.5% for H, S and ...

Predicting soil and forest floor carbon stocks in Western Anatolian Scots pine stands, Turkey

The objective of this study was to determine the models which estimate the soil plus forest floor carbon stocks (TC) using some stand, climate, and topographic parameters in Scots pine (Pinus sylvestris L.) stands. 40 sample plots were taken from sites differing in aspects, elevation, slope degree, slope position, and stand growth condition. TC of the stands, as dependent variables, and stand, climate, and topographic properties, as independent variables, were determined and evaluated with multiple regression analysis (MRA) and classification and regression tree technique (CART). In conclusion, TC were related with elevation, slope degree, slope position, aspect, actual evapotranspiration, mean annual precipitation, length of vegetation period, rainfall in the driest month, rainfall from June to September, stand volume, and site index. 91% of the variance in TC was explained with CART while 60% with MRA in a local scale. The results showed that soil and forest floor might become a carbon source where the rainfall decreased, but a sink, especially in productive sites where rainfall of the driest month increased. Key words: Carbon stocks, forest floor, regression tree, Scots pine, soil.

Identification of possible determinants of inpatient mortality using Classification and Regression Tree (CART) analysis in hospitalized oldest old patients

Current demographic trends suggest that there will be increasing numbers of older people in the future. Relatively little information is available regarding factors which influence mortality in the acutely unwell oldest old. This study uses the CART technique on data relating to the oldest old, to identify potential predictors of inpatient mortality in patients over 90 years old admitted acutely to the hospital due to various medical emergencies in two UK centers. The sample included 393 patients aged 90 years and older, with 67.5% females and 32.5% males and a mean age of 91.1 years. We aimed to generate hypotheses in order to identify potential acute illness prognostic indicators of inpatient mortality in this age group. The factors identified in this analysis which were associated with inpatient mortality in this patient population were raised serum urea concentration (>13.95mmol/L), low oxygen saturation levels (<94%), hyponatremia (<128mmol/L), and raised white cell count (>17×109/L). The predictability of using these cut off points in inpatient as well as early in-hospital death should be validated in future studies.

Predictors of preference for financial investment products using CART analysis

PurposeThe purpose of this paper is to illustrate the use of a post hoc predictive segmentation procedure to find out the variables that are the most important predictors of investor's preference for specific financial investment products.Design/methodology/approachThe study considers various demographic, socio‐economic and psychographic variables for the purpose of understanding the investor's preferences. Using a sample of individual investors (n=377), a classification and regression tree (CART) methodology was used to determine whether psychographic variables were better predictors than demographic and socio‐economic variables for understanding an individual investor's preference for the investment alternatives.FindingsThe results showed that psychographic variables emerged as the most important predictors in the case of investment products with greater degree of risk, and the demographic and socio‐economic variables emerged as the most important for the investment instruments with lesser degree of risk. However, when the sample was divided based on occupation profile (government and non‐government), for both the fixed returns based instruments and the non‐fixed instruments, psychographic variables emerged as the most important predictors.Practical implicationsThese results show the need for financial service providers to consider the psychographic variables along with demographic and socio‐economic variables, so as to better understand and advise the financial consumers. This would enable the financial service institutions to target their audience more sharply, so as to develop appropriate marketing strategies and further build the investor's trust.Originality/valueThis paper is a first of its kind to empirically identify the most important variable that determines the financial consumer's preference for investment products in India, using CART technique. This study contributes to furthering the understanding of investor behavior.

A CART technique to adjust production from longwall coal operations under ventilation constraints

Methane emissions in longwall coal mines can arise from a variety of geologic and production factors, where ventilation and degasification are primary control measures to prevent excessive methane levels. However, poor ventilation practices or inadequate ventilation may result in accumulation of dangerous methane–air mixtures. The need exists for a set of rules and a model to be used as guidelines to adjust coal production according to expected methane emissions and current ventilation conditions. In this paper, hierarchical classification and regression tree (CART) analyses are performed as nonparametric modeling efforts to predict methane emissions that can arise during extraction of a longwall panel. These emissions are predicted for a range of coal productivities while considering specific operational, panel design and geologic parameters such as gas content, proximate composition of coal, seam height, panel width, cut height, cut depth, and panel size. Analyses are conducted for longwall mines with and without degasification of the longwall panel. These models define a range of coal productivities that can be achieved without exceeding specified emissions rates under given operating and geological conditions. Finally, the technique was applied to longwall mines that operate with and without degasification system to demonstrate its use and predictive capability. The predicted results proved to be close to the actual measurements to estimate ventilation requirements. Thus, the CART-based model that is given in this paper can be used to predict methane emission rates and to adjust operation parameters under ventilation constrains in longwall mining.

Investigating Issues in Map Accuracy When Using an Object-Based Approach to Map Benthic Habitats

In a recent study, benthic habitat maps were created of the Texas Gulf Coast from digital aerial imagery. The images were classified using an object-based image analysis (OBIA) approach and a classification and regression tree (CART) technique. The map was manually edited, changing 26% of the polygons' labels. Accuracy assessments of the unedited map and the edited map revealed the two were not significantly different. The research in this paper evaluates why these maps may have similar accuracies. Our analyses indicate that the small segmentation scale parameter used over-segmented the imagery, reducing the effectiveness of the CART technique and editing.

Using Hotel Web Ratings Data: Understanding Customers' Overall Satisfaction Ratings

This article presents techniques that hotel managers can use to analyze online rating data and effectively allocate scarce resources to improve a hotel's overall customer ratings. We apply the Classification and Regression Tree (CART) technique to analyze online hotel customer component ratings and their relationship to overall hotel satisfaction ratings. We demonstrate that the CART technique shows how strongly component ratings associate with the overall hotel satisfaction rating. Although customer satisfaction ratings are important to customer decisions about which hotel to engage, we have found that little research has been done on this topic. We provide recommendations to hotel managers about how they can collect and analyze online ratings data to help improve the overall satisfaction of their customers.

Diagnosis of gastric cancer using near-infrared Raman spectroscopy and classification and regression tree techniques

The purpose of this study is to apply near-infrared (NIR) Raman spectroscopy and classification and regression tree (CART) techniques for identifying molecular changes of tissue associated with cancer transformation. A rapid-acquisition NIR Raman system is utilized for tissue Raman spectroscopic measurements at 785-nm excitation. 73 gastric tissue samples (55 normal, 18 cancer) from 53 patients are measured. The CART technique is introduced to develop effective diagnostic algorithms for classification of Raman spectra of different gastric tissues. 80% of the Raman dataset are randomly selected for spectral learning, while 20% of the dataset are reserved for validation. High-quality Raman spectra in the range of 800 to 1800 cm(-1) are acquired from gastric tissue within 5 s. The diagnostic sensitivity and specificity of the learning dataset are 90.2 and 95.7%; and the predictive sensitivity and specificity of the independent validation dataset are 88.9 and 92.9%, respectively, for separating cancer from normal. The tissue Raman peaks at 875 and 1745 cm(-1) are found to be two of the most significant features to discriminate gastric cancer from normal tissue. NIR Raman spectroscopy in conjunction with the CART technique has the potential to provide an effective and accurate diagnostic means for cancer detection in the gastric system.

Predicting materials properties and behavior using classification and regression trees

An investigation was conducted to evaluate the effectiveness of a non-parametric statistical methodology of classification and regression tree (CART) [L. Breiman, J.H. Friedman, R.A. Olshen, C.J. Stone, Classification and Regression Trees, Wadsworth Inc., California, 1984] as an alternative to the traditional parametric-based regression techniques in predicting materials properties and behavior. It has been demonstrated, with its application to a database on the creep rupture data of austenitic stainless steels, that the CART technique consistently outperforms the conventional curvilinear regression method in terms of the accuracy of prediction. Moreover, the results of the CART analysis provide an insight into the relationships and interactions between the materials variables and the insight will be beneficial to understanding materials behavior and useful in materials design.

Molecular abnormalities in the major psychiatric illnesses: Classification and Regression Tree (CRT) analysis of post-mortem prefrontal markers.

Post-mortem specimens from the Stanley Foundation Neuropathology Consortium, which contains matched samples from patients with schizophrenia, bipolar disorder, non-psychotic depression and normal controls (n = 15 per group), have been distributed to many research groups around the world. This paper provides a summary of abnormal markers found in prefrontal cortical areas from this collection between 1997 and 2001. With parametric analyses of variance of 102 separate data sets, 14 markers were abnormal in at least one disease. The markers pertained to a variety of neural systems and processes including neuronal plasticity, neurotransmission, signal transduction, inhibitory interneuron function and glial cells. The data sets were also examined using the non-parametric Classification and Regression Tree (CRT) technique for the four diagnostic groups and in pair-wise combinations. In contrast to the results obtained with analyses of variance, the CRT method identified a smaller set of nine markers that contributed maximally to the diagnostic classifications. Three of the nine markers observed with CRT overlapped with the ANOVA results. Six of the nine markers observed with the CRT technique pertained to aspects of glutamatergic, GABA-ergic, and dopaminergic neurotransmission.