J48 Decision Tree Research Articles

A machine leaning approach for abdominal aortic aneurysm severity assessment using geometric, biomechanical, and patient-specific historical clinical features.

Recent studies monitoring severity of abdominal aortic aneurysm (AAA) suggested that reliance on only the maximum transverse diameter ( D max ) may be insufficient to predict AAA rupture risk. Moreover, geometric indices, biomechanical parameters, material properties, and patient-specific historical data affect AAA morphology, indicating the need for an integrative approach that incorporates all factors for more accurate estimation of AAA severity. We implemented a machine learning algorithm using 45 features extracted from 66 patients. The model was generated using the J48 decision tree algorithm with the aim of maximizing model accuracy. Three different feature sets were used to assess the prediction rate: i) using D max as a single-feature set, ii) using a set of all features, and, lastly iii) using a feature set selected via the BestFirst feature selection algorithm. Our results indicate that BestFirst feature selection yielded the highest prediction accuracy. These results indicate that a combination of several specific parameters that comprehensively capture AAA behavior may enable a suitable assessment of AAA severity, suggesting the potential benefit of machine learning for this application.

Application of decision trees and fuzzy inference system for quality classification and modeling of black and green tea based on visual features

A fuzzified decision tree (DT)-based computer vision system was developed in this study for classification of common Iranian teas. Images of different tea categories, including five green tea and five black tea classes were captured using a CCD camera. In total, 83 image-based information (including 18 colour, 13 texture, and 52 wavelet-based features) were extracted from the images and introduced into DT classifier to distinguish between different tea categories. Reduced Error Pruning (REP) and J48 DTs were applied as classifier and the overall accuracy of both evaluated DTs was equal to 94.79% for differentiating all tea categories. However, these DTs had almost complex structures which made them unsuitable for developing a Mamdani fuzzy system, based on their structure. So, in order to simplify the structures of the DTs to be more suitable for developing fuzzy sets, the DTs were examined for categorizing black and green teas separately. The overall accuracy of J48 DT was 92.917% and 95.000% for classification of black and green teas, respectively and for REP tree the accuracy was equal to 90.83% and 97.08%, respectively. The REP tree structure was used to set the membership functions and rules of the fuzzy system because of the simpler structure of REP tree, even though J48 and REP trees had almost equal performances. It was concluded that the DT-based fuzzy logic system can be used effectively to classify different quality categories of tea, based on image extracted features.

A Comparative Analysis on Artificial Intelligence Techniques for Web Phishing Classification

Over the last few years, the web has been expanded to serve millions of users for various purposes all over the world. The web content filtering is essential to filter offensive, unwanted web content from web pages, reduced inappropriate content to prevent access to content which could compromise the network and spread malware. It also to tightened network security where web content filtering adds a much-needed layer of security to the network by blocking access to sites that raise an alarm. However, there are lack of comparison between classification techniques in previous studies in order to find the best classifier for the web page classification and the analysis related to it. Thus, the purpose of this study was to apply web page classification techniques and their performances is compared as it is the initial step in data mining before going to web filtering. In this project, three classifiers called Artificial Neural Network, J48 Decision Tree and Support Vector Machine were used to web phishing dataset in order to find the best possible classifier with small computational efforts that will give the best result in classifying the web page.

Analyzing Student Performance in Programming Education Using Classification Techniques

In this research, we aggregated students log data such as Class Test Score (CTS), Assignment Completed (ASC), Class Lab Work (CLW) and Class Attendance (CATT) from the Department of Mathematics, Computer Science Unit, Usmanu Danfodiyo University, Sokoto, Nigeria. Similarly, we employed data mining techniques such as ID3 & J48 Decision Tree Algorithms to analyze these data. We compared these algorithms on 239 classification instances. The experimental results show that the J48 algorithm has higher accuracy in the classification task compared to the ID3 algorithm. The important feature attributes such as Information Gain and Gain Ratio feature evaluators were also compared. Both the methods applied were able to rank search method and the experimental results confirmed that the two methods derived the same set of attributes with a slight deviation in the ranking. From the results analyzed, we discovered that 67.36 percent failed the course titled Introduction to Computer Programming, while 32.64 percent passed the course. Since the CATT has the highest gain value from our analysis; we concluded that it is largely responsible for the success or failure of the students.

Evaluating the usage of tree-based ensemble methods in groundwater spring potential mapping

The main goal of the study was to estimate the efficiency and predictive ability of a series of tree-based ensemble methods in groundwater spring potential mapping. The study introduces a novel hybrid integration approach based on J48 Decision Trees (J48), AdaBoost (AB), Bagging (Bag), RandomSubSpace (RS), Dagging (Dag) and Rotation Forest (RF) algorithms for constructing a groundwater spring potential map. The performance of the ensemble models were evaluated at the Wuqi County in Shaanxi Province, China. At first, a groundwater spring inventory map with 235 groundwater springs was constructed. The groundwater spring database was randomly divided into a training (70% of the total number of groundwater springs) and validating subset (remaining 30%). Secondly, sixteen groundwater spring related variables were selected and analyzed, such as elevation, slope, aspect, planform and profile curvature, curvature, topographic wetness index, stream transport and stream power indexes, distance from river network, lithology, soil and landuse cover, normalized difference vegetation index, mean annual rainfall and distance from road network. The contribution of each groundwater spring related variable so as to identify the most predictive was based on the correlation attribute evaluation method. The predictive performance of each model was evaluated by estimating the area under the receiver operating characteristic (ROC) curve (AUC) and several statistical indexes (accuracy, sensitivity, specificity). The outcomes revealed that all models had a good predictive performance, having AUC values greater than 0.74. Specifically, RF-J48 model appeared with the highest prediction capability (AUC = 0.797), followed by the AB-J48 model (0.793), the Bag-J48 (0.768), the RS-J48 (0.766) and the Dag-J48 (0.748), respectively. The study highlights that the proposed ensemble methodology is efficient and highly accurate appropriate for groundwater spring potential mapping.

Effectiveness assessment of Keras based deep learning with different robust optimization algorithms for shallow landslide susceptibility mapping at tropical area

This research aims at investigating the capability of Keras’s deep learning models with three robust optimization algorithms (stochastic gradient descent, root mean square propagation, and adaptive moment optimization) and two-loss functions for spatial modeling of landslide hazard at a regional scale. Shallow landslides at the Ha Long area (Vietnam) were selected as a case study. For this regard, set of ten influencing factors (slope, aspect, curvature, topographic wetness index, landuse, distance to road, distance to river, soil type, distance to fault, and lithology) and 193 landslide polygons were prepared to construct a Geographic Information System (GIS) database for the study area. Using the collected database, the DNN with its potential of realizing complex functional mapping hidden in the data is used to generalize a decision boundary that separates the learning space into two distinct categories: landslide (a positive class) and non-landslide (a negative class). Experimental results point out that the utilized the Keras’s deep learning model with the Adam optimization and the mean squared error lost function is the best with the prediction performance of 84.0%. The performance is better than those of the employed benchmark approaches of random forest, J48 decision tree, classification tree, and logistic model tree. We conclude that the Keras’s deep learning model is a new tool for shallow susceptibility mapping at landslide-prone areas.

Distinction of an Assortment of Deep Brain Stimulation Parameter Configurations for Treating Parkinson’s Disease Using Machine Learning with Quantification of Tremor Response through a Conformal Wearable and Wireless Inertial Sensor

Deep brain stimulation offers an advanced means of treating Parkinson’s disease in a patient specific context. However, a considerable challenge is the process of ascertaining an optimal parameter configuration. Imperative for the deep brain stimulation parameter optimization process is the quantification of response feedback. As a significant improvement to traditional ordinal scale techniques is the advent of wearable and wireless systems. Recently conformal wearable and wireless systems with a profile on the order of a bandage have been developed. Previous research endeavors have successfully differentiated between deep brain stimulation “On” and “Off” status through quantification using wearable and wireless inertial sensor systems. However, the opportunity exists to further evolve to an objectively quantified response to an assortment of parameter configurations, such as the variation of amplitude, for the deep brain stimulation system. Multiple deep brain stimulation amplitude settings are considered inclusive of “Off” status as a baseline, 1.0 mA, 2.5 mA, and 4.0 mA. The quantified response of this assortment of amplitude settings is acquired through a conformal wearable and wireless inertial sensor system and consolidated using Python software automation to a feature set amenable for machine learning. Five machine learning algorithms are evaluated: J48 decision tree, K-nearest neighbors, support vector machine, logistic regression, and random forest. The performance of these machine learning algorithms is established based on the classification accuracy to distinguish between the deep brain stimulation amplitude settings and the time to develop the machine learning model. The support vector machine achieves the greatest classification accuracy, which is the primary performance parameter, and K-nearest neighbors achieves considerable classification accuracy with minimal time to develop the machine learning model.

Exploiting Machine Learning Algorithms for Predicting Crash Injury Severity in Yemen: Hospital Case Study

This study focused on exploiting machine learning algorithms for classifying and predicting injury severity of vehicle crashes in Yemen. The primary objective is to assess the contribution of the leading causes of injury severity. The selected machine learning algorithms compared with traditional statistical methods. The filtrated second data collected within two months (August-October 2015) from the two main hospitals included 156 injured patients of vehicle crashes reported from 128 locations. The data classified into three categories of injury severity: Severe, Serious, and Minor. It balanced using a synthetic minority oversampling technique (SMOTE). Multinomial logit model (MNL) compared with five machine learning classifiers: Naive Bayes (NB), J48 Decision Tree, Random Forest (RF), Support Vector Machine (SVM), and Multilayer Perceptron (MLP). The results showed that most of machine learning-based algorithms performed well in predicting and classifying the severity of the traffic injury. Out of five classifiers, RF is the best classifier with 94.84% of accuracy. The characteristics of road type, total injured person, crash type, road user, transport way to the emergency department (ED), and accident action were the most critical factors in the severity of the traffic injury. Enhancing strategies for using roadway facilities may improve the safety of road users and regulations.

Computer-Aided Diagnosis of Chronic Kidney Disease in Developing Countries: A Comparative Analysis of Machine Learning Techniques

The high incidence and prevalence of chronic kidney disease (CKD), often caused by late diagnoses, is a critical public health problem, especially in developing countries such as Brazil. CKD treatment therapies, such as dialysis and kidney transplantation, increase the morbidity and mortality rates, besides the public health costs. This study analyses the usage of machine learning techniques to assist in the early diagnosis of CKD in developing countries. Qualitative and quantitative comparative analyses are, respectively, conducted using a systematic literature review and an experiment with machine learning techniques, with the k-fold cross-validation method based on the Weka © software and a CKD dataset. These analyses enable a discussion on the suitability of machine learning techniques for screening for CKD risk, focusing on low-income and hard-to-reach settings of developing countries, due to the specific problems faced by them, e.g., inadequate primary health care. The study results show that the J48 decision tree is a suitable machine learning technique for such screening in developing countries, due to the easy interpretation of its classification results, with 95.00% accuracy, reaching a nearly perfect agreement with an experienced nephrologist`s opinion. Conversely, random forest, naive Bayes, support vector machine, multilayer perceptron, and k-nearest neighbor techniques, respectively, yield 93.33%, 88.33%, 76.66%, 75.00%, and 71.67% accuracy, presenting at least moderate agreement with the nephrologist, at the cost of a more difficult interpretation of the classification results.

Evaluation of Prediction Algorithms in the Student Dropout Problem

University dropout is a growing problem which, in recent years, is using computer techniques to assist in the detection process. The paper presents the evaluation of some prediction algorithms to detect a student with a high possibility of scholar desertion. The approach uses real data from past scholar periods to create a dataset with different information of the students (i.e., personal, economic, and academic records). The algorithms selected in the experimental phase were: J48 decision tree, K-near neighbors, and support vector machine. We use two similarity metrics to split the dataset with cases with at least 80% of similarity to evaluate each case. We use the data from 2010 to 2016 with real students’ information to predict if there exists the possibility of a real academic dropout in one test for a period. The results show that the J48 algorithm reaches a better performance in both experiments. Besides, the tree generated for each student is taken as a path of attention, reaching around 88% of effectiveness. Finally, the conclusions argue the contributions of the paper and propose a future line of research.

Appling Data Mining Technique for Crime Prevention: The Case of Hossaena Town Police Office

The Law enforcement agencies like that of police today are faced with large volume of data that must be processed and transformed into useful information and hence data mining can greatly improve crime analysis and aid in reducing and preventing crime. The purpose of this study is to construct predictive models that could help in the effort of crime pattern analysis with the aim of supporting the crime prevention activities at the Hossaena town police office. For this study, a six-step hybrid knowledge discovery process model is followed, due to the nature of the problem and attributes in the dataset. The classification technique such as J48 decision tree and Naive Bayes used to build the models. Performance of the models is compared using accuracy, True Positive Rate, False Positives Rate, and the area under the Relative Optical character curve. J48 decision tree registers better performance with 96.34% accuracy. Lastly for extracting the knowledge the researcher develop the prototype for the user for support the decision which crime is assigned under the serious, medium or low for this purpose the researcher generate the prototypes. Keywords: Classification, Crime, Data Mining, Hybrid, WEKA DOI : 10.7176/CEIS/11-1-03 Publication date: January 31 st 2020

A Comparative Analysis of Data Mining Techniques on Breast Cancer Diagnosis Data using WEKA Toolbox

Breast cancer is considered the second most common cancer in women compared to all other cancers. It is fatal in less than half of all cases and is the main cause of mortality in women. It accounts for 16% of all cancer mortalities worldwide. Early diagnosis of breast cancer increases the chance of recovery. Data mining techniques can be utilized in the early diagnosis of breast cancer. In this paper, an academic experimental breast cancer dataset is used to perform a data mining practical experiment using the Waikato Environment for Knowledge Analysis (WEKA) tool. The WEKA Java application represents a rich resource for conducting performance metrics during the execution of experiments. Pre-processing and feature extraction are used to optimize the data. The classification process used in this study was summarized through thirteen experiments. Additionally, 10 experiments using various different classification algorithms were conducted. The introduced algorithms were: Naive Bayes, Logistic Regression, Lazy IBK (Instance-Bases learning with parameter K), Lazy Kstar, Lazy Locally Weighted Learner, Rules ZeroR, Decision Stump, Decision Trees J48, Random Forest and Random Trees. The process of producing a predictive model was automated with the use of classification accuracy. Further, several experiments on classification of Wisconsin Diagnostic Breast Cancer and Wisconsin Breast Cancer, were conducted to compare the success rates of the different methods. Results conclude that Lazy IBK classifier k-NN can achieve 98% accuracy among other classifiers. The main advantages of the study were the compactness of using 13 different data mining models and 10 different performance measurements, and plotting figures of classifications errors.

A research on influence factors of user's tie strength integrated with topic orientation

This paper studies the influence factors of tie strength between users in online social networks while considering the effect of topic orientation on user behaviour. We compare various models to identify the optimal classification model of tie strength. Furthermore, we explore the importance of different attributes including the users' different behaviours under different topic orientation on tie strength between users. The research finds that the most effective classification model is decision tree J48. According to the importance ranking of attributes, it is also found that the factor of frequency of interaction has the most prominent ability to distinguish tie strength in general. In general, users' different interaction behaviours with diverse topic orientation have varying abilities to distinguish the tie strength. The results of this study are helpful to support the further development of research in tie strength calculation.

The Multi-Class Classification for the First Six Surats of the Holy Quran

The Holy Quran is one of the holy books revealed to the prophet Muhammad in the form of separate verses. These verses were written on tree leaves, stones, and bones during his life; as such, they were not arranged or grouped into one book until later. There is no intelligent system that is able to distinguish the verses of Quran chapters automatically. Accordingly, in this study we propose a model that can recognize and categorize Quran verses automatically and conclusion the essential features through Quran chapters classification for the first six Surat of the Holy Quran chapters, based on machine learning techniques. The classification of the Quran verses into chapters using machine learning classifiers is considered an intelligent task. Classification algorithms like Naive Bayes, SVM, KNN, and decision tree J48 help to classify texts into categories or classes. The target of this research is using machine learning algorithms for the text classification of the Holy Quran verses. As the Quran texts consists of 114 chapters, we are only working with the first six chapters. In this paper, we build a multi-class classification model for the chapter names of the Quranic verses using Support Vector Classifier (SVC) and GaussianNB. The results show the best overall accuracy is 80% for the SVC and 60% for the Gaussian Naive Bayes.

A Model for Prediction of Kidney Cancer Using Data Analytics Technique

Across the globe, kidney cancer and other cancerous diseases has been a threat to human lives. The incidence and mortality rate represent a significant and growing threat to both developed and developing countries especially in Africa, where most cancers are diagnosed at an advanced stage. This typically contributes to its complications and high rate of mortality, and has been attributed to limited awareness of early signs and symptoms of the disease, lack of detective mechanism and inaccessible cancer care in our health care centres. To preclude the harm and mortality caused by the disease, an intelligent mechanism for early prediction and prognosis of the syndrome is vital. However, early detection and prognosis requires an accurate information and analytic procedure that will assist and equip the health-care providers/public with the skills to identify early the indicators of the disease. Efforts in this work, produced a model for early prediction of kidney cancer using data analytic approach. Dataset and reports pertaining to the disease were acquired from selected private and public hospitals in fifty-two (52) selected LGA in Nigeria. A two-layered classifier system consisting of Artificial Neural Networks (ANN) and Decision Tree (DT) designed for the work was successfully employed in the model building. Waikato Environment for Knowledge Analysis (WEKA) platform was used for the experiment. The performance of the classifiers considered was compared using standard metrics of accuracy and time taken as benchmark. Experimental results show that the J48 decision tree algorithm outperform all other algorithms in the classifier family with correctly classified instances of 74.7%, F-Measure of 0.614, TP rate of 0.747, FP rate of 0.135, precision and recall of 0.687 and 0.714 respectively. It took the algorithm, 0.03 seconds to build the model. The performance of this algorithm proved its suitability as a valuable tool for the research purpose. The model will in no small measure support the efforts of the national health scheme in preventing the disease mortality rate.

Prediction of Wheat Rust Diseases Using Data Mining Application

Stem rust caused by Puccinia graminis f.sp. tritici and stripe rust caused by Puccinia striiformis are the most destructive wheat rust diseases when environment conditions are favorable in regions where wheat crops are grown. An early prediction mechanism can play a great role in forecasting the occurrence of the rust disease. It assists proactive control and early decision making. However, in the absence of prediction mechanism of wheat rust disease hurried wheat production yield loss more. Hence, to overcome these issues, this study was conducted to develop wheat stripe rust and stem rust diseases prediction model using data mining application. The meteorological and disease data for the year 2010 to 2019 of national meteorological agency southern Oromia region Bale Robe service center and Oromia Seed enterprise Sinana farm unit II were obtained and used for the study. Two bread wheat varieties, namely Kakaba and Danda’a were involved in the study. Fist, daily meteorological data mean values of three-day consecutive interval were computed. Then, mean value of meteorological data has been integrated with disease incidence and severity date observed during the most critical infection period (mid-august to November-30) was used to develop the model. WEKA software machine learning tool with J48 decision tree algorithm was used for data preprocess and experiments. It is open source software containing various machine learning algorithms for data mining tasks. The study results showed that, stripe rust predictive model trained and tested with using training set test option achieved accuracy 75.70% and stem rust predictive model achieved accuracy 90.045% with cross-validation fold 10 test option. The results found were much promised to forecast occurrence of wheat stripe and stem rust disease.

Comparative Study on Tool Fault Diagnosis Methods Using Vibration Signals and Cutting Force Signals by Machine Learning Technique

The state of cutting tool determines the quality of surface produced on the machined parts. A faulty tool produces poor surface, inaccurate geometry and non-economic production. Thus, it is necessary to monitor tool condition for a machining process to have superior quality and economic production. In the present study, fault classification of single point cutting tool for hard turning has been carried out by employing machine learning technique. Cutting force and vibration signals were acquired to monitor tool condition during machining. A set of four tooling conditions namely healthy, worn flank, broken insert and extended tool overhang have been considered for the study. The machine learning technique was applied to both vibration and cutting force signals. Discrete wavelet features of the signals have been extracted using discrete wavelet transformation (DWT). This transformation represents a large dataset into approximation coefficients which contain the most useful information of the dataset. Significant features, among features extracted, were selected using J48 decision tree technique. Classification of tool conditions was carried out using Naïve Bayes algorithm. A 10 fold cross validation was incorporated to test the validity of classifier. A comparison of performance of classifier was made between cutting force and vibration signal to choose the best signal acquisition method in classifying tool fault conditions using machine learning technique.

Comparison of Accuracy Between Two Methods: Naїve Bayes Algorithm and Decision Tree-J48 to Predict The Stock Price of Pt Astra International tbk Using Data From Indonesia Stock Exchange

The ability to predict the stock prices is very important for market players, whether individual or organizational investors. The market players needs to know how to predict, that will help them in their decision making process, whether to buy or to sell its shares, so that it can maximize profits and reduce potential losses due to mistakes in decision making. In accordance to this, the authors conducted a study that aimed to analyze and to compare the accuracy of two (2) methods that is used to predict the stock prices, namely: the Naїve Bayes Method and the Decision Tree-J48 Method. The amount of data used in this study were 1,195 stock datas ofPT Astra International Tbk, issued by the IDX, by the period of January 1, 2013 to November 30, 2017.This study uses 7 attributes, namely: Previews, High, Low, Close, Volume, Value, and Frequency. By using the WEKA application the result shows that, the accuracy of the Naïve Bayes Method using 20% of testing data, is 92.0502%, the precision value is 0.920 and the value of recall is 0.961, while the accuracy of the Decision Tree J-48 method, using 20% of testing data, is 98.7448%, with precision value of 0.989 and the value of recall of0.997.Through this results, it can be concluded that the decision tree J-48 algorithm has a better accuracy results compared to the Naive Bayes algorithm in predicting the stock price of PT. Astra Internasional Tbk.

Sentiment classification of Chinese cosmetic reviews based on integration of collocations and concepts

PurposeThis paper aims to propose a novel approach which integrates collocations and domain concepts for Chinese cosmetic word of mouth (WOM) sentiment classification. Most sentiment analysis works by collecting sentiment scores from each unigram or bigram. However, not every unigram or bigram in a WOM document contains sentiments. Chinese collocations consist of the main sentiments of WOM. This paper reduces the complexity of the document dimensionality and makes an improvement for sentiment classification.Design/methodology/approachThis paper builds two contextual lexicons for feature words and sentiment words, respectively. Based on these contextual lexicons, this paper uses the techniques of associated rules and mutual information to build possible Chinese collocation sets. This paper applies preference vector modelling as the vector representation approach to catch the relationship between Chinese collocations and their associated concepts.FindingsThis paper compares the proposed preference vector models with benchmarks, using three classification techniques (i.e. support vector machine, J48 decision tree and multilayer perceptron). According to the experimental results, the proposed models outperform all benchmarks evaluated by the criterion of accuracy.Originality/valueThis paper focuses on Chinese collocations and proposes a novel research approach for sentiment classification. The Chinese collocations used in this paper are adaptable to the content and domains. Finally, this paper integrates collocations with the preference vector modelling approach, which not only achieves a better sentiment classification performance for Chinese WOM documents but also avoids the curse of dimensionality.

Classification of web services using data mining algorithms and improved learning model

As per the global digital report, 52.9% of the world population is using the internet, and 42% of the world population is actively using e-commerce, banking, and other online applications. Web services are software components accessed using networked communications and provide services to end users. Software developers provide a high quality of web service. To meet the demands of user requirements, it is necessary for a developer to ensure quality architecture and quality of services. To meet the demands of user measure service quality by the ranking of web services, in this paper, we analyzed QWS dataset and found important parameters are best practices, successability, availability, response time, reliability and throughput, and compliance. We have used various data mining techniques and conducted experiments to classify QWS data set into four categorical values as class1, 2, 3, and 4. The results are compared with various techniques random forest, artificial neural network, J48 decision tree, extreme gradient boosting, K-nearest neighbor, and support vector machine. Multiple classifiers analyzed, and it was observed that the classifier technique eXtreme gradient boosting got the maximum accuracy of 98.44%, and random forest got the accuracy of 98.13%. In future, we can extend the quality of web service for mixed attributes.