Classification Tree Technique Research Articles

Intrusion Detection System using Datamining Based Enhanced Framework

With the significant increase in the use of computers over the network and the development of applications on different platforms, the focus is on network security. The identification of multiple attacks is actually an important element of network security. The role of the IDS is to track and prevent unauthorized use or damage to network resources and systems. An intrusion detection system using Datamining Based Enhanced Framework (DEF) is presented in this paper. The model is assisted by the K-mean Clustering and Decision Tree (DT) classification techniques in which genetic algorithms (GA) for clusters, max runs and confidence can be used. The experimental results shows the promising outcome of the proposed Datamining Based Enhanced Framework (DEF).

Field study on the factors affecting egg weight loss, early embryonic mortality, hatchability, and chick mortality with the use of classification tree technique.

The aim of this research was to study the factors affecting hatchability, percentage of egg weight loss during incubation, early embryonic mortality, and chick mortality in the first seven days of life in the field. This study was performed on 20,817,600 hatching eggs originating from 7 different breeder flocks. The factors responsible for variability of investigated traits were the following: genotype (ROSS 308, ROSS PM3), hen age (25 to 30, 31 to 45, 46 to 50, and 51 to 60 wk), egg storage time (0 to 13 d), and setter and hatcher type (Digital, Airstreamer, Vision, Focus, Biostreamer), which were determined using the classification tree technique. Statistical calculations were performed using R software, version 3.4.3. Moreover, 2-way analysis of variance followed by Duncan's multiple comparison test was performed. The breeder flock age and egg storage time were the most important factors responsible for variability in the percentage of fertile hatchability; however, setter and hatcher type also affected the level of this trait. The highest hatchability was observed for eggs laid by hens aged 31 to 50 wk and stored up to 6 d. Genotype was the most important determinant of the percentage of egg weight loss and that Ross 308 eggs lose less weight when compared to Ross PM3 ones. Egg storage time was the most important factor, which affected early embryonic mortality. The present study has shown that the 2 main factors, i.e., breeder flock age and egg storage time, that affected hatchability have also influenced chick mortality. Moreover, it was found that both hatcher and setter type affected chick mortality in the first 7 d of life. The importance of this research lies in the fact that besides showing the most important factors affecting hatching process, we were also able to suggest how to adjust the management decisions at commercial hatcheries in order to increase production results in different egg sets, which vary in respect of breeder flock age, egg storage time, setter, and hatcher type and genotype.

Breast Cancer Tumor Categorization using Logistic Regression, Decision Tree and Random Forest Classification Techniques

Breast Cancer Tumor Categorization using Logistic Regression, Decision Tree and Random Forest Classification Techniques

Decision tree for modeling survival data with competing risks

This work considers decision tree for modeling survival data with competing risks. A Survival Classification and Regression Tree (SCART) technique is proposed for analysing survival data by modifying classification and regression tree (CART) algorithm to handle censored data for both regression and classification problems. Different performance measures for regression and classification tree are proposed. Model validation is done by two different cross-validation methods. Two real life data sets are analyzed for illustration. It is found that the proposed method improve upon the existing classical method for analysis of survival data with competing risks.

Investigation of helmet use behavior of motorcyclists and effectiveness of enforcement campaign using CART approach

Use of the helmet is beneficial in reducing the severity of injuries and avoiding fatalities for motorcyclists, therefore, legislation exists in almost all countries. In practice, the situation is different regarding the helmet use, especially in the developing countries. This paper investigates the helmet use behavior of motorcyclists in Karachi, Pakistan. It determines the significant factors affecting the helmet use in Karachi (Pakistan) and recommending effective campaign measures to promote helmet use. It is vital in relation to the exponential growth of motorcycle users and poor enforcement of traffic rules. Repeated cross-sectional data collected before (n = 226) and after (n = 277) the helmet use enforcement campaign is analyzed using univariate and non-parametric classification and regression tree (CART) techniques. A significant number of motorcyclists do not hold a driving license and CART analysis highlighted the significance of this variable along with exposure to the road environment (measured as daily usage of motorcycle) to explain helmet use. Campaign effects are found temporary, therefore, serious efforts are required to design coherent and structured awareness and enforcement programs.

Discrimination of rice genotypes using field spectroradiometry

Spectral discrimination of rice genotypes was investigated using canopy reflectance in the range of 350 to 2500 nm. The pre-processed reflectance spectra were statistically analysed using one-way analysis of variance (ANOVA) followed by classification and regression tree (CART) technique to find significantly sensitive wavelengths for discrimination. The CART was able to select seventeen wavelengths (4 in visible, 5 in near-infra-red and 8 in shortwave infra-red region) well distributed over the entire spectrum. The spectral separability between each pair of rice genotypes at the selected wavebands was quantified using Jeffries–Matusita (J–M) distance analysis. The J–M distance analysis taking 91 pairs of genotypes showed that all the pairs were separable. This result was further validated by quadratic discriminant analysis (QDA) with an overall accuracy of 98%. The variation in biophysical and biochemical attributes of genotypes has been captured through differential spectral reflectance at selected wavebands which could make the discrimination possible.

POTENTIAL DISTRIBUTION MODELLING AND MAPPING USING RANDOM FOREST METHOD: AN EXAMPLE OF YUKARIGÖKDERE DISTRICT

Reasons such as climate abnormalities and anthropogenic effects on ecosystems threat the existence of plant and animal species. In this case, it is important to be more effective and functional planning studies for ecosystems and target species. In order to achieve the mentioned effectiveness and functionality, creating of potential distribution maps for the target species are of crucial importance. There are many methods (discriminant analysis, logistic regression analysis, generalized addicted model, classification tree technique, maximum entropy, the genetic algorithm for rule-set prediction etc.) used in the process of creating potential distribution maps. These methods process with different algorithms and can produce different results. The Random Forest (RF) method, which is one of the mentioned methods and is frequently preferred in recent years, is used in this study. RF method was run via RStudio. In the present, AUC value of the potential distribution model obtained by RF method was determined to be 0.978. The variables creating to the model are respectively elevation (elvn), topographic position index (tpi), bedrock (rock), radiation index (ri) and slope (slope).

Prediction of strontium bromide laser efficiency using cluster and decision tree analysis

Subject of investigation is a new high-powered strontium bromide (SrBr2) vapor laser emitting in multiline region of wavelengths. The laser is an alternative to the atom strontium lasers and electron free lasers, especially at the line 6.45 μm which line is used in surgery for medical processing of biological tissues and bones with minimal damage. In this paper the experimental data from measurements of operational and output characteristics of the laser are statistically processed by means of cluster analysis and tree-based regression techniques. The aim is to extract the more important relationships and dependences from the available data which influence the increase of the overall laser efficiency. There are constructed and analyzed a set of cluster models. It is shown by using different cluster methods that the seven investigated operational characteristics (laser tube diameter, length, supplied electrical power, and others) and laser efficiency are combined in 2 clusters. By the built regression tree models using Classification and Regression Trees (CART) technique there are obtained dependences to predict the values of efficiency, and especially the maximum efficiency with over 95% accuracy.

Artificial neural networks and decision tree classification for predicting soil drainage classes in Denmark

Soil drainage constitutes a substantial factor affecting plant growth and various biophysical processes, such as nutrient cycling and greenhouse gas fluxes. Consequently, soil drainage maps represent crucial tools for crop, forest and environmental management purposes. As extensive field surveys are time- and resource-consuming, alternative spatial modelling techniques have been previously applied for predicting soil drainage classes. The present study assessed the use of Artificial Neural Networks (ANN) for mapping soil drainage classes in Denmark and compared it to a Decision Tree Classification (DTC) technique. 1702 soil observations and 31 environmental variables, including soil and terrain parameters, and spectral indices derived from satellite images, were utilized as input data. Based on a 33% holdback validation dataset, the best performing ANN and DTC models yielded overall accuracy values of 54 and 52%, respectively. DTC models benefited from the use of all variables, but ANN models performed better after variable selection. Notably, ANN and DTC model performances were comparable although differential costs for misclassification were only implemented for DTC modelling. Nevertheless, both methods produced predictive drainage maps in accordance with one another and demonstrated promising classification abilities over a large study area (c. 43,000km2).

Decision trees in epidemiological research

BackgroundIn many studies, it is of interest to identify population subgroups that are relatively homogeneous with respect to an outcome. The nature of these subgroups can provide insight into effect mechanisms and suggest targets for tailored interventions. However, identifying relevant subgroups can be challenging with standard statistical methods.Main textWe review the literature on decision trees, a family of techniques for partitioning the population, on the basis of covariates, into distinct subgroups who share similar values of an outcome variable. We compare two decision tree methods, the popular Classification and Regression tree (CART) technique and the newer Conditional Inference tree (CTree) technique, assessing their performance in a simulation study and using data from the Box Lunch Study, a randomized controlled trial of a portion size intervention. Both CART and CTree identify homogeneous population subgroups and offer improved prediction accuracy relative to regression-based approaches when subgroups are truly present in the data. An important distinction between CART and CTree is that the latter uses a formal statistical hypothesis testing framework in building decision trees, which simplifies the process of identifying and interpreting the final tree model. We also introduce a novel way to visualize the subgroups defined by decision trees. Our novel graphical visualization provides a more scientifically meaningful characterization of the subgroups identified by decision trees.ConclusionsDecision trees are a useful tool for identifying homogeneous subgroups defined by combinations of individual characteristics. While all decision tree techniques generate subgroups, we advocate the use of the newer CTree technique due to its simplicity and ease of interpretation.

Twitter’s Sentiment Analysis on Gsm Services using Multinomial Naïve Bayes

Telecommunication users are rapidly growing each year. As people keep demanding a better service level of Short Message Service (SMS), telephone or data use, service providers compete to attract their customer, while customer feedbacks in some platforms, for example Twitter, are their souce of information. Multinomial Naive Bayes Tree, adapted from the method of Multinomial Naive Bayes and Decision Tree, is one technique in data mining used to classify the raw data or feedback from customers.Multinomial Naive Bayes method used specifically addressing frequency in the text of the sentence or document. Documents used in this study are comments of Twitter users on the GSM telecommunications provider in Indonesia.This research employed Multinomial Naive Bayes Tree classification technique to categorize customers sentiment opinion towards telecommunication providers in Indonesia. Sentiment analysis only included the class of positive, negative and neutral. This research generated a Decision Tree roots in the feature aktif in which the probability of the feature aktif was from positive class in Multinomial Naive Bayes method. The evaluation showed that the highest accuracy of classification using Multinomial Naive Bayes Tree (MNBTree) method was 16.26% using 145 features. Moreover, the Multinomial Naive Bayes (MNB) yielded the highest accuracy of 73,15% by using all dataset of 1665 features. The expected benefits in this research are that the Indonesian telecommunications provider can evaluate the performance and services to reach customer satisfaction of various needs.

Impact of an Onsite Clinic on Utilization of Preventive Services.

To assess impact of an onsite clinic on healthcare utilization of preventive services for employees of a public university and their dependents. Descriptive statistics, logistic regression and classification tree techniques were used to assess health claim data to identify changes in patterns of healthcare utilization and factors impacting usage of onsite clinic. Utilization of preventive services significantly increased for women and men employees by 9% and 14% one year after implementation of the onsite clinic. Hourly-paid employees, employees without diabetes, employees with spouse opt out or no coverage were more likely to go to the onsite clinic. Adapted framework for assessing performance of onsite clinics based on usage of health informatics would help to identify health utilization patterns and interaction between onsite clinic and offsite health providers.

Infants and the decision to provide ongoing child welfare services

BackgroundInfants are the most likely recipients of child welfare services; however, little is known about infants and families who come into contact with the child welfare system and factors that are associated with service provision. Investigations involving infants and their families present an unparalleled opportunity for the child welfare sector to enhance infants’ safety and well-being through early identification, referral and intervention. Understanding how the child welfare system responds to the unique needs of infants and caregivers is critical to developing appropriate practice and policy responses within the child welfare sector and across other allied sectors. This study examines maltreatment-related investigations in Ontario involving children under the age of one to identify which factors are most influential to predicting service provision at the conclusion of a child welfare investigation.MethodsA secondary analysis of the fifth cycle of the Ontario Incidence Study of Reported Child Abuse and Neglect (OIS) for 2013 was conducted. The OIS is a cross-sectional child welfare study that is conducted every 5 years. The most influential factors that were associated with the decision to transfer a case to ongoing services were explored through a multivariate tree-classification technique, Chi square automatic interaction detection.ResultsThere were an estimated 7915 maltreatment-related investigations involving infants in 2013. At least one caregiver risk factor was identified in approximately three-quarters (74%) of investigations involving infants. In the majority of investigations (57%), at least one referral for specialized services was provided. Primary caregiver with few social supports was the most highly significant predictor of the decision to provide ongoing child welfare services. Primary caregiver risk factors were predominant in this model. The analysis identified subgroups of investigations involving infants for which the likelihood of being transferred to ongoing services ranged from approximately 11–97%.ConclusionCaregivers of infants are struggling with numerous challenges that can adversely compromise their ability to meet the unique developmental needs of their infant. The findings underscore the importance of community and social supports in decision-making.

Use of the Decision Tree Technique to Estimate Sugarcane Productivity Under Edaphoclimatic Conditions

A number of biometric evaluations are performed during harvest for measuring the growth and development of the sugarcane crop. From these evaluations, hundreds of data values are generated, containing certain information on the productivity of the culture in that crop and edaphoclimatic region. Accordingly, the objective of this work was to identify, using a decision tree classification technique, the biometric attribute having the greatest effect on the productivity of the plant cane in different planting configurations and edaphoclimatic conditions. To accomplish this, data were evaluated from four experiments with sugarcane, located within the Sao Paulo municipalities of Teodoro Sampaio, Guaira, Iracemapolis, and Lencois Paulista. The classification model was generated using the decision tree technique, a type of intuitive learning that creates a hypothesis based on particular instances that results in general conclusions. The decision trees applied to the data of the four sites showed that the population of plants per hectare has the highest information gain (split attribute) on the class attribute (productivity). Using the “Chi-square” method of attribute selection, the population of plants per hectare was observed to have the largest correlation with the final productivity of the culture. Therefore, the decision tree indicates that the attribute “plant population per area” should be used as the method to evaluate the productive potential of the culture during its growth cycle. It has the best correlation with the final productivity of the crop, in addition to being an attribute easy to measure in the field.

Applying object-based image analysis and knowledge-based classification to ADS-40 digital aerial photographs to facilitate complex forest land cover classification

In general, considerable human and material resources are required for performing a forest inventory survey. Using remote sensing technologies to save forest inventory costs has thus become an important topic in forest inventory-related studies. Leica ADS-40 digital aerial photographs feature advantages such as high spatial resolution, high radiometric resolution, and a wealth of spectral information. As a result, they have been widely used to perform forest inventories. We classified ADS-40 digital aerial photographs according to the complex forest land cover types listed in the Fourth Forest Resource Survey in an effort to establish a classification method for categorizing ADS-40 digital aerial photographs. Subsequently, we classified the images using the knowledge-based classification method in combination with object-based analysis techniques, decision tree classification techniques, classification parameters such as object texture, shape, and spectral characteristics, a class-based classification method, and geographic information system mapping information. Finally, the results were compared with manually interpreted aerial photographs. Images were classified using a hierarchical classification method comprised of four classification levels (levels 1 to 4). The classification overall accuracy (OA) of levels 1 to 4 is within a range of 64.29% to 98.50%. The final result comparisons showed that the proposed classification method achieved an OA of 78.20% and a kappa coefficient of 0.7597. On the basis of the image classification results, classification errors occurred mostly in images of sunlit crowns because the image values for individual trees varied. Such a variance was caused by the crown structure and the incident angle of the sun. These errors lowered image classification accuracy and warrant further studies. This study corroborates the high feasibility for mapping complex forest land cover types using ADS-40 digital aerial photographs.

Wykorzystanie technik data mining w analizowaniu czynnik�w wplywajacych na reaktywnosc kr�w podczas doju

Motor activity of 158 Polish Holstein-Friesian cows was evaluated 5 times (before and during milking in a DeLaval 2*10 milking parlour) for both the morning and evening milking, on a 5-point scale, according to the method of Budzynska et al. (2007). The statistical analysis used multiple logistic regression and classification trees (Enterprise Miner 7.1 software which comes in with SAS package). In the evaluation of motor activity, cows that were among the first ten to enter the milking parlour were more often given a score of 3 points before (11.5%) and during milking (23.5%) compared to the other cows. Cows’ activity tended to decrease (both before and during milking) with advancing lactation. The cows’ reduced activity was accompanied by shorter teat cup attachment times and lower milk yields. The criteria calculated for the quality of models based on classification tree technique as well as logistic regression showed that similar variables were responsible for the reactivity of cows before milking (teat cup attachment time, day of lactation, number of lactation, side of the milking parlour) and during milking (day of lactation, side of the milking parlour, morning or evening milking, milk yield, number of lactation). At the same time, the applied methods showed that the determinants of the cow reactivity trait are highly complex. This complexity may be well explained using the classification tree technique.

Are Nomothetic or Ideographic Approaches Superior in Predicting Daily Exercise Behaviors?

The understanding of how stress influences health behavior can provide insights into developing healthy lifestyle interventions. This understanding is traditionally attained through observational studies that examine associations at a population level. This nomothetic approach, however, is fundamentally limited by the fact that the environment- person milieu that constitutes stress exposure and experience can vary substantially between individuals, and the modifiable elements of these exposures and experiences are individual-specific. With recent advances in smartphone and sensing technologies, it is now possible to conduct idiographic assessment in users' own environment, leveraging the full-range observations of actions and experiences that result in differential response to naturally occurring events. The aim of this paper is to explore the hypothesis that an ideographic N-of-1 model can better capture an individual's stress- behavior pathway (or the lack thereof) and provide useful person-specific predictors of exercise behavior. This paper used the data collected in an observational study in 79 participants who were followed for up to a 1-year period, wherein their physical activity was continuously and objectively monitored by actigraphy and their stress experience was recorded via ecological momentary assessment on a mobile app. In addition, our analyses considered exogenous and environmental variables retrieved from public archive such as day in a week, daylight time, temperature and precipitation. Leveraging the multiple data sources, we developed prediction algorithms for exercise behavior using random forest and classification tree techniques using a nomothetic approach and an N-of-1 approach. The two approaches were compared based on classification errors in predicting personalized exercise behavior. Eight factors were selected by random forest for the nomothetic decision model, which was used to predict whether a participant would exercise on a particular day. The predictors included previous exercise behavior, emotional factors (e.g., midday stress), external factors such as weather (e.g., temperature), and self-determination factors (e.g., expectation of exercise). The nomothetic model yielded an average classification error of 36%. The ideographic N-of-1 models used on average about two predictors for each individual, and had an average classification error of 25%, which represented an improvement of 11 percentage points. Compared to the traditional one-size-fits-all, nomothetic model that generalizes population-evidence for individuals, the proposed N-of-1 model can better capture the individual difference in their stressbehavior pathways. In this paper, we demonstrate it is feasible to perform personalized exercise behavior prediction, mainly made possible by mobile health technology and machine learning analytics.

Are Charter Value and Supervision Aligned? A Segmentation Analysis

Previous work suggests that the charter value hypothesis is theoretically grounded and empirically supported, but not universally. Accordingly, this paper aims to perform an analysis of the relations among charter value, risk taking, and supervision that take into account the relations’ complexity. Specifically, using the CAMELS rating system as a general framework for supervision, we carry out a segmentation analysis of how charter value relates to risk and supervision by means of classification and regression tree techniques. The sample covers the period 2005–2014 and consists of listed banks in countries that were members of the eurozone when it came into existence, along with Greece. Our results reflect the complexity of the relations among charter value, supervision, and risk. Indeed, supervision and charter value seem aligned regarding only some types of risk, but the relation is not constant through time or across countries.

TBM performance estimation using a classification and regression tree (CART) technique

With widespread increasing applications of mechanized tunneling in almost all ground conditions, prediction of tunnel boring machine (TBM) performance is required for time planning, cost control and choice of excavation method in order to make tunneling economical. Penetration rate is a principal measure of full-face TBM performance and is used to evaluate the feasibility of the machine and predict the advance rate of an excavation. In this study, a database of actual machine performance from T05 and T06 tunnels of the deep tunnel sewerage system (DTSS) project in Singapore which include: rock mass uniaxial compressive strength, brittleness index (Bi), volumetric joint account (Jv), joint orientation (Jo), TBM specifications and corresponding TBM performance has been compiled. Then, for prediction of specific rock mass boreability index (SRMBI), two different models including classification and regression tree (CART) analysis and multivariate regression analysis (MVRA) have been developed. As statistical indices, correlation coefficient (R2), root mean square error (RMSE) and variance accounted for (VAF) were used to evaluate the efficiency of the developed models for determining the SRMBI of TBMs. According to the obtained results, it was observed that the performance of the CART model is better than the MVRA.

Correlation between lymph node count and survival and a reappraisal of lymph node ratio as a predictor of survival in gastric cancer: A multi-institutional cohort study

PurposeThe purpose of this study is to evaluate the correlation between lymph node count (LNC) and survival and to evaluate whether lymph node ratio (LNR) which is related to LNC is a better predictor of survival for gastric cancer than the N category of UICC/AJCC through a multi-institutional cohort study. MethodsThe study cohort included 3284 patients from eight institutions. Lower and upper quartiles of LNC were used for comparisons. The cut-off values (0, 0.06, 0.27, and 0.49) for the LNR categories were based on Classification and Regression Trees techniques. Akaike information criteria (AIC) for Cox regression models was used to evaluate goodness of fit between competing predictor variables (LNR vs. N category). ResultsThe 5-year disease-specific survival (DSS) rates of lower and upper quartiles of LNC were 82.2% and 84.8%. In the subgroup analysis of pN category, the upper quartile of LNC showed better survival than the lower quartile in pN2, pN3a, and pN3b subgroups. Regarding LNR, 5-year DSS of LNR 0, 0–0.06, 0.06–0.27, 0.27–0.49, and >0.49 was 95.3%, 88.7%, 70.6%, 42.7%, and 17.2% respectively. Multivariate analysis showed that pT, pN, LNR, residual tumor status, distant metastasis, and tumor differentiation significantly affected survival. The analysis also confirmed superiority of LNR compared with N category in the AIC analysis. ConclusionHigher LNC correlated with better survival in patients with pN2, pN3a, and pN3b gastric cancer. Our data indicate that LNR is a better predictor of survival than N category of UICC/AJCC.