Data Mining Classification Methods Research Articles

A fast neighborhood classifier based on hash bucket with application to medical diagnosis

In the medical diagnosis, expensive costs will increase significantly with the increment of medical information, and they can be reduced by data mining methods. The neighborhood classifier, as one of the extensions of the neighborhood rough set, has become an intuitive and effective classification method in data mining. However, there are still some defects which limit its performance. On the one hand, most existing neighborhood classifiers suffer from high computation complexity to obtain the neighborhood of samples; on the other hand, the difference between samples in the neighborhood is ignored when classifying samples, which diminishes the classification ability of the model. Therefore, in this paper, hash buckets and distance voting rule are introduced to solve the above problems, and a fast neighborhood classifier based on hash bucket (FNC-HB) is proposed. First, all samples are mapped to corresponding buckets through the hash function. Next, for any test sample, a bucket-based adaptive neighborhood classification radius is defined, in which the artificial parameter has been eliminated. Then, to avoid the indiscernibility of traditional voting rule when predicting labels of the test sample, a new rule called distance voting rule is defined. Finally, experimental results show that FNC-HB has better classification performance and computation efficiency, which is feasible and effective.

Heart Disease Prediction System Based on Multiple Feature Selection Algorithm with Ensemble Classifier

In today's world, deaths from heart disease have become a significant problem. Nearly one person dies from heart disease every minute. In most situations, the diagnosis of cardiac disease is based on a complicated mix of clinical and pathological data. This complexity leads to expensive medical costs, which have an impact on the quality of medical care. The intention of the work is to predict the heart disease based on medical records. The proposed method employs data mining classification methods like KNN, SVM, and Decision Tree to find out knowledge in a heart disease dataset, which is obtained from UCI Machine learning repository. The input dataset consists of 282 observations with 75 attributes. The feature selection of input dataset is done using Multiple Feature Selection Algorithm (MFSA), which uses Auto-Correlation and Information gain to find the best set of features and improves the performance of the classifiers. The results are obtained from classification methods with and without using Feature Selection in the form of accuracy. Based on the experimental analysis, classification with Feature Selection method outperformed classification without Feature Selection method.

Nursing Diagnosis of Urology Operating Room Based on New Association Classification Algorithm.

Due to the rapid development of medical engineering, massive amounts of data are recorded and preserved by various medical instruments. Therefore, finding relationships among data and summarizing clinical manifestations are of great significance to the diagnosis, treatment, and medical research of various diseases. The key to studying the nursing diagnosis support system, particularly in the urological operating room, is to select an effective classification algorithm, which is suitable for the characteristics of urological diseases. Initially, we have analyzed characteristics of urological diseases through medical data mining. Secondly, based on the traditional data mining classification method and urological disease diagnosis research, we have introduced the urological disease experimental source dataset and analyzed characteristics of the disease. Furthermore, classification algorithm and steps were introduced such as decision tree (including ID3, C4.5), Bayesian classification, BP neural network, and association rule classification algorithms. These algorithms are used to make relevant comparative experiments on the urological disease dataset. Finally, based on the diagnosis of urological diseases, a new association classification algorithm (ACCF), which is based on frequent closed item sets, is proposed along with suitable explanation. In order to verify the operational capabilities, the proposed algorithms are implemented in C++ and compared with the classification effect of traditional association classification algorithms and data mining methods. Both theoretical analysis and experiment results show that the proposed algorithm has resolved various deficiencies of the existing data mining algorithms and equally improved the accuracy of urological disease classification and prediction.

Intelligent Data Analysis using Optimized Support Vector Machine Based Data Mining Approach for Tourism Industry

Data analysis involves the deployment of sophisticated approaches from data mining methods, information theory, and artificial intelligence in various fields like tourism, hospitality, and so on for the extraction of knowledge from the gathered and preprocessed data. In tourism, pattern analysis or data analysis using classification is significant for finding the patterns that represent new and potentially useful information or knowledge about the destination and other data. Several data mining techniques are introduced for the classification of data or patterns. However, overfitting, less accuracy, local minima, sensitive to noise are the drawbacks in some existing data mining classification methods. To overcome these challenges, Support vector machine with Red deer optimization (SVM-RDO) based data mining strategy is proposed in this article. Extended Kalman filter (EKF) is utilized in the first phase, i.e., data cleaning to remove the noise and missing values from the input data. Mantaray foraging algorithm (MaFA) is used in the data selection phase, in which the significant data are selected for the further process to reduce the computational complexity. The final phase is the classification, in which SVM-RDO is proposed to access the useful pattern from the selected data. PYTHON is the implementation tool used for the experiment of the proposed model. The experimental analysis is done to show the efficacy of the proposed work. From the experimental results, the proposed SVM-RDO achieved better accuracy, precision, recall, and F1 score than the existing methods for the tourism dataset. Thus, it is showed the effectiveness of the proposed SVM-RDO for pattern analysis.

Implementation of Logistic Regression Classification Algorithm and Support Vector Machine for Credit Eligibility Prediction

Credit is a provision of money or bills that can be equated with it, the provision of loans or credit. A good credit analysis is very necessary, because it is one of the most important processes in the form of an investigation regarding the smooth or substandard credit repayments. The stages of identifying and predicting customers properly and correctly can be done before the loan process. This is done by examining the historical data of the customer's loan. At this time this activity is an effort made by the banking industry in dealing with credit risk problems. In this research, researchers will apply several data mining classification methods, including Logistic Regression algorithms and Support Vector Machines to predict creditworthiness. The dataset used 481 record motorized vehicle loan data, both problematic and non-problematic. The input variables in this study consisted of thirteen variables, including marital status, number of dependents, age, residence status, home ownership, occupation, employment status, company status, income, down payment, education, length of stay, and housing conditions. From the results of research and testing, the performance of the Logistic Regression model for predicting creditworthiness provided an accuracy rate of 94.81% with an area under the curve (AUC) value of 0.987. While the performance of the Support Vector Machine model provides an accuracy of 94.19% with an area under the curve (AUC) value of 0.978. Based on the T-Test test, the Logistic Regression method has the same performance compared to the Support Vector Machine.

Attribute Weighted Na飗e Bayes Classifier

The naïve Bayes classifier is one of the commonly used data mining methods for classification. Despite its simplicity, naïve Bayes is effective and computationally efficient. Although the strong attribute independence assumption in the naïve Bayes classifier makes it a tractable method for learning, this assumption may not hold in real-world applications. Many enhancements to the basic algorithm have been proposed in order to alleviate the violation of attribute independence assumption. While these methods improve the classification performance, they do not necessarily retain the mathematical structure of the naïve Bayes model and some at the expense of computational time. One approach to reduce the naïveté of the classifier is to incorporate attribute weights in the conditional probability. In this paper, we proposed a method to incorporate attribute weights to naïve Bayes. To evaluate the performance of our method, we used the public benchmark datasets. We compared our method with the standard naïve Bayes and baseline attribute weighting methods. Experimental results show that our method to incorporate attribute weights improves the classification performance compared to both standard naïve Bayes and baseline attribute weighting methods in terms of classification accuracy and F1, especially when the independence assumption is strongly violated, which was validated using the Chi-square test of independence.

IMPLEMENTASI ALGORITMA DECISION TREE UNTUK KLASIFIKASI PRODUK LARIS

Decision Tree C4.5 algorithm is an algorithm that can be used to make a decision tree. Decision tree (Decision Tree) is one method that is quite easily interpreted by humans. However, this algorithm has never been tested for product classification using private data (stock data and sales of goods at PT Cipta Karya Gorontalo). Therefore this study aims to test the accuracy of C4.5 in classifying best-selling products (private data). As a result of the evaluation of product classification models using Decision Tree C4.5 obtained from this study amounted to 90% and AUC value of 0.709 where this value is included in the Good Classification. It can be used as a data mining classification method Decision Tree C4.5 algorithm is accurate in classifying hot-selling products.
 
 Keywords— Decision Tree, C4.5, Classification, Best-Selling Product

STRATEGI PEMASARAN PRODUK INDUSTRI KREATIF MENGGUNAKAN ALGORITMA K-MEANS CLUSTERING BERBASIS PARTICLE SWARM OPTIMIZATION

The existence of abundant UMKM data sources can be used to dig up information. Classification is one of the techniques to explore hidden data owned by data mining. Data mining classification methods, one of which is the Support Vector Machine (SVM) algorithm. The SVM algorithm has proven better results than the KKN, Decision Tree and Linear Regression algorithms. In the classification process, the accuracy and time efficiency results obtained are very important. So optimization is needed in order to increase accuracy and time efficiency during the classification process. The optimization of the SVM algorithm was carried out using the K-Means algorithm for the clustering and continuous process on UMKM data and the feature selection process using Particle Swarm Optimization (PSO). This paper aims to optimize the accuracy of the data in the form of type of business, business and turnover. From the results of the discussion of the SVM method using K-Means and PSO, it gives an average accuracy of 55% but 0.12% lower than SVM just using PSO. Keywords: UMKM, Clustering, K-Means, SVM, PSO

Palm oil extraction rate prediction based on the fruit ripeness levels using C4.5 algorithm

Oil palm plantations are one of the main keys in supporting Indonesia’s economic growth. The rising consumption needs for palm oil products make it necessary to carry out data mining activities to increase CPO production. The maturity factor of palm fruit dramatically affects the quality of the oil extraction content (CPO yield) produced. This study aims to investigate the effect of fruit ripeness on the yield of CPO by using a data mining classification method with a decision tree. The algorithm used to generate decision tree classification is the C4.5 algorithm. The implementation of the C4.5 algorithm in the study was carried out using the Rapid Miner Studio 5.2 tools. The results shows that the yield of CPO is influenced by the attributes of the condition of the long and ripe fruit, the condition of the long and overripe fruit, the normal condition of the fruit and the age of 3-6 years and the condition of the fruit of normal and age of 7-10 years. Decision tree C4.5 algorithm generates 8 rules with 4 rules showing a high production value, which means that the four rules affect the yield of CPO.

Plant Leaf Disease Detection by Using Different Classification Techniques: Comparative

One of the main factors that assist to increase the growth of any country is Agriculture. The detection of diseases from plant leaf images is one of the most important fields of agricultural research. To identify disease factors, this field requires a reliable prediction approach. Data Mining (DM) is the process of analyzing data from different aspects and summarizing it into valuable information. It helps users to categorize and identify relationships between data from various dimensions. As there are many plants on the farm, detecting and classifying the diseases of each plant on the farm is extremely difficult for the human eye. And diagnosing each plant is very critical since these diseases may spread. DM classification is an important method that has a wide range of applications. It classifies each item in a set of data into one of a set of predefined classes. In this paper, a comparison of different DM classification methods such as Naive Bayes, Decision trees, SVM, and Random Forest algorithms has been illustrated by using of Weka Tool.

PREDICTION FOR COOPERATIVE CREDIT ELIGIBILITY USING DATA MINING CLASSIFICATION WITH C4.5 ALGORITHM

BMT Artha Mandiri is a cooperative that provides savings and loans services. In providing credit, BMT Artha Mandiri still uses the manual method, namely by looking at the ledger and history of each customer, to find out whether the applicant is worthy or not worthy of credit so that it is not effective and efficient. The purpose of this research is to make an application that can predict whether a prospective customer is eligible or not to be given credit. Predictions are made using the data mining classification method, namely the C4.5 algorithm based on the supporting data each customer has to classify which factors have the most influence on the level of credit payments in the cooperative. In a built application, the C4.5 algorithm produces a decision tree that is easy to interpret based on the existing variables. In the application, there are features that can be used to make decisions about customers who will apply for credit at the cooperative. The blackbox test results on the application show that the application has been able to run as expected, while the results of the algorithm test also show that the application has been able to implement the C4.5 algorithm correctly. In addition, the results of testing for accuracy show that the maximum average value of Accuracy is 79.19%.

Implementasi Algoritma Naïve Bayes dan K-Nearest Neighbor Dalam Menentukan Tingkat Keberhasilan Immunotherapy Untuk Pengobatan Penyakit Kanker Kulit

Cancer is a disease that grows in the skin tissue where this condition is characterized by changes in the skin, such as the appearance of lumps, spots, or moles with abnormal sizes, one of the causes of skin cancer is exposure to ultraviolet rays from the sun. One of the treatments for skin cancer is immunotherapy, the immunotherapy method is the treatment of disease by activating or suppressing the immune system in the body. In this study, a comparison with data mining methods for classification was carried out, namely Naïve Bayes and K-Nearest Neighbor to predict the success rate of immunotherapy in curing skin cancer. In the testing process, the researcher uses the Weka application to process data and conduct tests. The results of the tests that have been carried out show that the K-Nearest Neighbor model has the best accuracy value of 91.1111%. while Naïve Bayes obtained a smaller accuracy value, namely 82.2222%. From the test results, it can be concluded that the K-Nearest Neighbor method has better accuracy in determining the success rate of immunotherapy.

Science, Technology and Innovation Policy Indicators and Comparisons of Countries through a Hybrid Model of Data Mining and MCDM Methods

Science, technology and innovation (STI) policies are of great importance for countries to reach their sustainable development goals. Numerous global databases have many indicators that measure and compare the performance of STI policies of countries. However, many problems arise regarding how to identify, classify and systematically analyze these indicators in order to measure, monitor and improve the performance of STI. The study includes a literature review on global problems and new trends in STI policies, while mentioning the necessity of an internationally comparable STI indicator set, current STI indicator studies and efforts, and studies for each continent. In light of these, all the indicators selected are introduced in detail. The strengths and weaknesses of the countries in the study in terms of evaluation indicator values are indicated. After determining the indicator weights objectively with the entropy method, 40 countries are compared with TOPSIS, VIKOR, PROMETHEE I-II, ARAS, COPRAS, MULTIMOORA, ELECTRE, SAW and MAUT methods. In addition, countries that show similarities with each other are evaluated by cluster analysis, which is one of the data mining classification methods. This study offers a new and original approach with MCDM methods on this subject. Considering all the results obtained in the study together, these rankings are compared among themselves and with the rankings specified in the Global Innovation (2019) and Global Competitiveness (2019) indices, and it is seen that the results are consistent. In addition, it is possible to update and publish this study every year with updated data.

Prediksi Kebutuhan Alat Kesehatan Rumah Sakit Menggunakan Metode Algoritma Regression Linier dan Naïve Bayes

 The hospital is a health service institution for the community with its own characteristics that require a variety of resources in carrying out its activities. One of the most important is health equipment. Medical devices are supporting aspects that support the implementation of health services. PALI Regional General Hospital is a Type D hospital, which needs to manage its medical devices. Kepmenkes Regulation No. 004 / MENKES / SK / 1/2003 concerning health policy and strategy on decentralization in the health sector states that one of the strategic objectives is the effort to organize health management in the decentralization era is to develop sub-systems of maintenance and optimization of utilization of health facilities and equipment. The amount of medical device data can only be estimated from the many or at least the available medical devices (stock), because the needs of each year are different. This results in not all the needs of medical devices being met and often additional stocks occur while the amount of APBD has been divided for each institution. So to anticipate this it is necessary to predict the need for medical devices in PALI District Hospital. If the status of medical device needs can be predicted early, the hospital can minimize data redundancy (repetition of data) and information can be up to date (update). In this study, the authors will predict medical devices in Pali District Hospital using the classification method in data mining based on the Algortima Linear Regression model to get the most accurate test results.Keywords : Prediction, Medical Devences, Linear Regression Algorithm, Naive Bayes

Hybrid Radial Basis Function Neural Networks for Urban Traffic Signal Control

In this study, a real-world isolated signalized intersection with a fixed-time signal control system is considered. The signal timing plans are arranged regardless of the traffic density, and these plans cause delays in vehicle queues. To increase the efficiency of the intersection, an adaptive traffic signal control system is proposed to manage the intersection. To find the appropriate adaptive green times for each lane, simulations are performed by traffic simulation software using vehicle arrivals and other information about vehicle movements gathered from the real-world intersection. Then, a hybrid radial basis function neural network is developed to forecast the adaptive green times, which is trained and tested with historical arrivals and simulation results. The performance of the proposed network is compared with well-known data mining classification methods, such as support vector regression, k-nearest neighbors, decision tree, random forest, and multilayer perceptron methods, by different evaluation parameters. The comparison results provide that the developed radial basis function neural network outperforms other classification methods and can be successfully used for forecasting adaptive green times as an alternative to complex unsupervised classification methods.

The Determination Analysis Of Telecommunications Customers Potential Cross-Selling With Classification Naive Bayes And C4.5

Every company has various marketing strategies. Marketing strategies are offered to sell the company’s products to benefit if the company introduces new products. However, too many offers to customers that are not true, will only make marketing inefficient and ineffective. Data mining as a way to find patterns and relationships in data can be used to make valid predictions. To simplify the marketing strategy of PT. TELKOM then classifies the data that already exists in PT. TELKOM Jakarta. Telkom customers have the potential to become customers of new Indihome products, so marketing is carried out by PT. TELKOM has become more effective and efficient. By using data mining classification methods, namely the Naive Bayes Classifier algorithm and the C4.5 algorithm, patterns and relationships are obtained to simplify the marketing strategy of PT. TELKOM where in previous research, the results of classification of data mining research models with the Naive Bayes Classifier algorithm have an accuracy value of 85.08% and AUC 0.841 while in this study the C4.5 algorithm has an accuracy value of 88.61% and AUC 0.870. C4.5 is a model with good accuracy for customer classification data that has the potential for more effective and efficient in cross-selling marketing strategy..

Rice Leaf Disease Image Classifications Using KNN Based On GLCM Feature Extraction

The disease that often occurs in rice leaves which causes crop failure. At present the attack of pests or leaf diseases in countries where the majority of rice as the main food of the community, especially in Indonesia is increasing. Early detection mechanism to reduce the risk of crop failure is very necessary. Data mining classification methods for the past few years have been very popular to be used in detecting the classification of rice leaf disease. Our paper uses 120 images of rice leaf disease from the UCI repository. The purpose of this study was to determine how to classify images of rice leaf disease consisting of three diseases namely Bacterial leaf blight, Brown spot, and Leaf smut. The study proposes the GLCM method as feature extraction for text analysis, with five feature values consisting of contrast, energy, entropy, homogeneity, and correlation. KNN (K-Nearest Neighbor) algorithm is used for the classification of rice leaf disease, by finding the maximum k value from the experiment k value 1 to 20. The results of our experiments show that the value of k = 11 has the highest accuracy value compared to other k values of 65.83% and kappa 0.485.

Penerapan Algoritma Decision Tree Untuk Memprediksi Penerima Bantuan Keluarga Harapan

The Family of Hope Program (PKH) is a poverty reduction program in the education and public health aspects provided by the government either directly or indirectly. The government continues to make efforts in order to educate the community through social assistance programs to tackle the poor. In order to create a smart society, the government should make programs that are empowering so that people can solve their own problems. There are many in Indonesia who receive the family hope program (PKH), one of which is in the Cirebon district. Problems often occur with the empowerment assistance program from the government, one of which is the PKH assistance, which still does not target the residents who receive the assistance. The emergence of this problem, due to the ineffective data verification in determining which citizens are entitled to receive PKH assistance, this has resulted in many very poor people who do not receive PKH assistance and those classified as capable are still given PKH assistance. Therefore, it is necessary to conduct a study of PKH beneficiary data, so that the results of the analysis can be used as a reference for whether or not residents are eligible to receive PKH assistance. The research that was conducted to predict the data of recipients of the expected family assistance using the data mining classification method using the C4.5 algorithm. The results of the data mining process are used as evaluation material for the government. After testing with the C4.5 algorithm, the test results for the best parameter of the C4.5 algorithm are criterion = accuracy, confidence = 0.25 and a minimum gain = 0.1 to produce an accuracy value of 98.30%

Data Mining Model For Designing Diagnostic Applications Inflammatory Liver Disease

Hepatitis is an infectious disease that is a public health problem that affects morbidity, mortality, public health status, life expectancy, and other socio-economic impacts. Early diagnosis of hepatitis is very important so that it can be treated and treated quickly. In this study, the authors will apply and compare several data mining classification methods, including the C4.5 algorithm, Naïve Bayes, and k-Nearest Neighbor to diagnose hepatitis, then compare which of the three methods is the most accurate. Based on the results of measuring the performance of the three models using the Cross Validation, Confusion Matrix and ROC Curve methods, it is known that the C4.5 method is the best method with an accuracy of 70.99% and an under the curva (AUC) value of 0.950, then the k-Nearest Neighbor method with accuracy of 67.19% and the value under the curve (AUC) 0.873, then the naïve Bayes method with an accuracy rate of 66.14% and a value under the curve (AUC) of 0.742.
 

Classification of Relationship between Alumni Competence and Work Using Decision Tree Method

This study discusses the classification of the relationship between personal competence with work on the tracer study data of the PLN Institute of Technology. The results of the data from the alumni survey will be processed using a decision tree method that aims to provide information related to the relationship between personal competencies which have a very close relationship, close enough and not close to the work that the alumni get and become an evaluation material for the campus in the future. Where the decision tree method is one of the data mining classification methods that provides a tree structure that will serve as a decision tree. The tracer study data used from the graduation years of 2015, 2016, 2017, 2018 and 2019 are processed based on the year. And consists of 13 competency attributes and one competency that is used as a label or target. Based on the application of this method each year of graduation has a different level of accuracy based on the state of each data, for tracer study data 2015 graduation year accuracy of 56.03%, 2016 graduation data accuracy of 56.03%, while for tracer data the 2017 graduation year study was 78.90% accuracy, for the 2018 graduation study tracer data the accuracy was 52.94% and for the 2019 graduation tracer study data the accuracy was 40.12%.