Search

Students graduating on time are an essential indicator for a higher education institution in supporting campus accreditation. Several factors cause students to graduate on time, namely the origin of the previous student's school and student interest. This study aims to predict student graduation based on the head of the last student's school and student interest so that higher education institutions can get the basis for decisions that will be taken in the future. The method used in analyzing student data and supporting criteria for predicting student graduation is the C4.5 algorithm. Then for the decision tree classifier, this research uses data mining. [ABSTRACT FROM AUTHOR]

Predicting annual workers income requires ones to deep dive into several factors. Factors that majorly being discuss were age, gender, education and occupation. On the other hand, there are other factors that may affect the annual workers income where it yet to be discussed. The traditional way of predicting the annual workers income was multiple linear regression. This parametric approach requires assumptions to be fulfilled and this will actions is a time-consuming activity. Data mining approach in predicting the workers income is important to understand on how the economy and compensation work in the United States. Machine learning will cover all aspect without needing to fulfil certain assumptions as compared to traditional method. Hence, the best way to predict the worker's income in the United States is the best using machine learning and concurrently solve the SDG 8: Decent Work & Economic Growth aspect. The dataset used in this study is acquired from Kaggle website. At first, features weight using filter method (Weight by Information Gain, Weight by Information gain Ratio and Weight by Chi – Squared Statistics) were taken to identify the influential factors towards annual workers' income. The three different methods employed in the model to predict worker income are logistic regression, decision trees, and artificial neural networks. The second goal is to contrast the effectiveness of worker income prediction using under sampling and oversampling techniques. The results show that, with the exception of decision tree, oversampling strategy provides the best performance of prediction model when compared to under sampling technique. Since under sampling techniques randomly delete observations when there is a chance that such observations could be significant to the data and have an impact on the prediction model, oversampling techniques perform better than under sampling techniques. The third goal is to identify the most effective classification model for predicting worker's income. The oversampling strategy with backward selection represents the best model when applying the Logistic Regression model. Additionally, the optimal model for Decision Trees is the backward selection with under sampling strategy. The best model criterion for artificial neural networks is the oversampling method via backward selection. Data mining approach in predicting the workers income is important to understand on how the economy and compensation work in the United States. [ABSTRACT FROM AUTHOR]

The improving of a company's earnings is considered as one of the important Challenges that facing these companies in terms of the amounts of their manufactured products. Therefore, it is very essential to determine which units of production that gives the reasonable profit. In this paper, data Mining (DM) and decision tree were used to define the products quantities of the Iraqi carpet company in the year 2016 that led to the acceptable profit using decision tree algorithm. This said algorithm is determined the product amounts that can gave the required profit namely; low permeable carpet, high embroidered blanket and high woven blanket with their quantities138, 1172, and 5025 respectively. In order to check the validity of the acceptable profit, these products were tested by the Naive Bayes algorithm which check their probabilities as probability of P (yes) and P (no) got 0.0208 and 0 respectively. [ABSTRACT FROM AUTHOR]

Data mining helps a company to make decisions. Companies can collect and analyze data and then use that data for decision-making. This case study focuses on examining how a courier company can improve its employee satisfaction index. Due to the lack of employee satisfaction in the company, we use three data mining methods to help the company solve its problems. There are three methods used: (1) Decision Tree to get a pattern and classification, (2) Association Rules to find correlated attributes, and (3) K-Means Clustering to find groups in data. By using these three methods, the data obtained can be processed and analyzed using the Rapid Miner tools. This study uses employee data in a courier company, where the employees who work have their respective satisfaction level factors. Several attributes are used to complete the required data, such as emp_id, age, dept, location, education, recruitment_type, job_level, rating, work_location, award, certification, salary, satisfaction. This study shows that employee satisfaction is not necessarily maximal, several factors that can affect employee satisfaction include awards, salaries, education, and certification. Classification technique has accuracy rate of 86.36%, Association technique has a level of confidence that varies from 0.548 to 0.960 and Clustering technique has a Davies Bouldin index of 1,092 with 10 clusters. [ABSTRACT FROM AUTHOR]

Breast cancer starts in the breast and develops into a tumour as a result of aberrant cell proliferation. One out of every four cancer cases in women worldwide is breast cancer. As the disease advances, survival becomes more difficult, and the woman in Indiawith more than half of breast cancer cases is diagnosed with advanced stages of breast cancer. Three data mining approaches are used in this study to build a model for early breast cancer diagnosis: Decision Tree, Support Vector Machine, and Artificial Neural Network, and their performance are compared to find the best method for breast cancer prediction. These approaches were tested using the Breast Cancer Wisconsin Original Dataset, which is available in the UCI Machine Learning Repository. The K-Nearest Neighbor approach is used to pre-process the data in this study so that missing attribute values can be filled in the dataset. According to the findings and comparison of outcomes, the algorithm Artificial Neural Network outperforms the other two methods for breast cancer prediction. [ABSTRACT FROM AUTHOR]

Across the globe, diabetes is recognized as one of the many causes of deaths, especially in Third World countries as there is a lack of treatment for diabetes, especially in the early stages. In study, the presence of diabetes will be classified within the community, thus contributing to the existing technology within the healthcare system. Our discovery can help doctors to predict the existence of diabetes accurately and alert patients to seek early treatments. Four data mining algorithms were used within this study which consists of both single and ensemble classifiers. The two single classifiers are decision tree, and logistic regression classifier while the ensemble classifiers are random forest, and stacking. These classifiers are chosen as they are efficient and high in performance. This research uses the PIMA diabetes dataset as it can be obtained by the general public. The stratify cross-validation is used to ensure the efficiency of the models. Ensemble classifiers show better or similar testing results compared to single classifiers. From data visualisation, two important features are discovered. [ABSTRACT FROM AUTHOR]

The C4.5 algorithm still has weaknesses in predicting or classifying data if the amount is large. It is necessary to improve the performance of the C4.5 algorithm with the selected split attribute using application of the average gain value to perform the classification. The C4.5 algorithm is one of the Decision Tree methods in the classification process using entropy. The result of the classification obtained from the analysis can be a classification of 8 student data from 100 student data that is tested to produce information on Sanguine, Choleric, Melancholy, and Phlegmatic. From the result of the Decision Tree classification algorithm C4.5 has an accuracy rate of 86.36% with an application error rate or error of 13.64%. [ABSTRACT FROM AUTHOR]

The K-Means method is the most popular and often used function in Data Mining in data grouping. One of the shortcomings of the K-Means method is the determination of the number of clusters that have not been optimal. This study will develop a combination of grouping datasets using the K-Means Clustering Method with the Davies Bouldin Index (DBI) for determining the most optimal number of clusters. The result of clustering is then classified using the Decision Tree Method with parameter optimization to produce the best performance indicated by the highest accuracy, recall and precision values. The processed dataset is public data from UCI Machine Learning Repository which is divided into training data and testing data. From performance tests conducted for the number of clusters 3 to 10, it is known that the most optimal number of clusters is 3 with the smallest Index Davies Bouldin value of 0.626 where cluster 0 has 490 members, cluster 1 has 124 members and cluster 2 has 197 members. Testing classification methods with K-Cross Validation for training data resulted in an accuracy value of 98.06 with a precision value of 97.64 and a recall value of 97.49. For data testing generate accuracy value 97.12 with precision value 97.10 and recall value 96.05. Because the accuracy value of>90, the classification in this study is excellent clasification. [ABSTRACT FROM AUTHOR]

Customer churn has become a big problem for telecommunication companies. Preventive efforts are needed by predicting the value of churn in the future. This study uses data mining techniques with decision tree algorithms to predict customer churn in one of Indonesian Telecommunication companies. The best decision tree model has parameters of criterion information gain with a minimal gain = 0.01 and a max depth = 6. This decision tree model has an accuracy value of 78.28% with 19,6% customer churn rate. Based on this model, customers of this company tend to have voluntary churns. Some important factors that affect customer churn are type of contract, number of monthly downloads, tenure, customer satisfaction value, and add on. The type of contract has the highest impact on the customer churn in this company. Based on the results, the company is suggested to promote a retention program based in order to decrease customer churn rate. [ABSTRACT FROM AUTHOR]

The classification algorithm's goal is to built a model that maximizes the accuracy of the number of correct predictions, although the completeness of the model plays an important role in many application areas. Ant Colony Optimization (ACO) is relatively simple to realize the behavior of ant colonies, and they cooperate with each other to achieve the goal from nest to food source. A system capable of executing a search to discover the optimum answer to an optimization issue with a vast search space is referred to as a colony generation system. Classification by applying the ACO algorithm in data mining has the advantage of searching with flexible values and value combinations. One of the many benefits that can be applied using ACO is to build a decision tree. As a model representation, the decision tree is easy to understand and can be represented in the form of a graph. By using the modified decision tree using ACO, the result of using the pruning technique is 76.1%. [ABSTRACT FROM AUTHOR]

This paper focuses on finding the best classification algorithm model in the case study of student performance prediction and comparing the algorithm performance before and after using the normalization method. To achieve this goal, this study uses data mining classification techniques to analyse student performance at Vocational High School in 2020-2021. The steps of the research carried out include: data collection, data pre-processing, build algorithm models without using normalization and with using normalization, and final step are comparing algorithm performance before and after using normalization. The algorithms that will be used include: Random Forest, Decision Tree, Logistic Regression, SVM, Naive Bayes, and KNN. While the normalization methods used are Standard Scaler, Min-Max Scaler, and Robust Scaler. The result of this research is that the normalization method is able to significantly increase the accuracy of the model. Based on the tests and evaluations carried out, the normalization method using the Min-Max Scaler has the biggest impact in improving the overall model performance and the algorithm with the best performance is Random Forest. This paper reviews the effect of the normalization method to improve algorithm performance in predicting student performance, where based on previous research no one has used the normalization method to gain accuracy of the model which actually has a considerable impact on gaining accuracy of the model. [ABSTRACT FROM AUTHOR]

This study aims to measure and analyze customer satisfaction with the services provided by OV Agency agents. Methos The data used in the study are questionnaires filled by 100 random customers from OV Agency. The study used the Fuzzy-ServQual method and data mining with the Classification-Decision Tree algorithm. Fuzzy-ServQual is used to measure attributes that need to be improved based on the calculation of GAP values with the addition of fuzzy methods to eliminate the obscurity obtained from the results of the questionnaire. The Data Mining Classification Method was used to determine hidden patterns by a decision tree. The results of the calculation of the overall GAP value on the Fuzzy-ServQual method are -0.008 and there are 10 attributes that require improvement. Decision trees are using the Rapid Minner Studio application. Based on the decision tree obtained, if then rule can be obtained as a pattern of satisfaction and dissatisfaction. [ABSTRACT FROM AUTHOR]

Heart Disease is one of the diseases that cause deaths in large numbers, so predicting these diseases is one of the most necessary tasks in the field of heart to reduce the number of deaths by providing health care to patients, so the process of forecasting these diseases aids in reducing disease-related risks and alerting patients to them. In this paper, Machine Learning (ML) algorithms were applied to a set of medical data to help diagnose these diseases and predict the chances of infection by using different data mining techniques like Decision Tree (DT), Naive Bayes (NB), Random Forest (RF), and Logistic Regression (LR). The dataset used 615 samples of medical data from the Kaggle Repository. After a comparison between these algorithms, the RF algorithm had the highest accuracy of 97.40% then LR with 90.24%, NB at 87.1545 %, and finally, DT with the low accuracy of 72.3577 %. It also used other evaluation matrices such as precision, recall, and f-measure for each algorithm, then compare accuracy founding with another paper using the same dataset and algorithm. [ABSTRACT FROM AUTHOR]

Data mining has a role to produce a decision. In implementing data mining, we use the Rapid Miner application. The Rapid Miner application has a function to analyze and implement data mining methods on the data set which is the basis of this research. In the data set we obtained, we have several attributes to complete data needs, namely Invoice_ID, and Ratings etc. The problem to be investigated in this study is the analysis of the transaction history on the supermarket sales data set using the data mining method, namely classification with the Decision Tree, clustering with K-Means, and association with the Association Rules Algorithm. This study aims to explain and analyze the supermarket transaction history, based on the information obtained from the analysis of data mining methods. The research method used is qualitative methods in the form of case studies. Based on research that has been conducted on supermarket sales data sets, information is obtained from several data mining methods and there is a lack of accuracy of information on the association method. So that in the supermarket sales data set that we have analyzed, there are several things that need to be improved. [ABSTRACT FROM AUTHOR]

Data mining has many algorithms; one of the most frequently used is the decision tree algorithm. This article described the result of the research of a data mining approach that used the decision tree algorithm; it aimed to create and provide a recommendation system for university students to choose and decide the topic for their final projects. The final project (so-called as thesis) is one of the graduation requirements for undergraduate students; many university students are confused in choosing and deciding the right topic for their thesis, and this results in not completing it in the required time. The impact of overdue thesis is the extension of their period of studies. In this study, data mining was carried out based on 9 attribute data sets which were reduced from 21 derived factors. The data processing of the research has used the free software so-called KNIME. It was started with attribute correlation testing; it was continued by formulating KNIME workflow and resulted in the decision tree. Then, the tree diagrams were translated into the 65 IF-THEN rules. Based on these 65 IF-THEN rules, this study developed a recommender system using C# language; the recommendation system was tested on 45 students which gave its accuracy more than 70%. [ABSTRACT FROM AUTHOR]

Due to the environmental condition people face various health hazards and it affects their living habits. So predicting the disease at earlier stage becomes important. As a specific symptom can be responsible for various diseases, it is time consuming to predict the perfect one for the doctors. Data mining plays a significant role in this most challenging task i.e. the accurate prediction of diseases. Due to increase in data growth in every year in healthcare field the precise analysis of data helps patients to take early care. Using the disease data, hidden pattern information was found by data mining in the enormous amount of medical data. Here in this study based on the symptoms general disease prediction techniques have been proposed. For the disease prediction, K-Nearest Neighbor (KNN), decision tree and random forest machine learning algorithm were used. [ABSTRACT FROM AUTHOR]

Churn prediction methods are widely used to anticipate customer churn from services provided by a company for some reasons. This study aims to develop an optimal churn prediction model based on customer data from a telecommunication company in Indonesia. The model development and evaluation processes are performed by following the Cross-Industry Standard Process for Data Mining (CRISP-DM), which consist of business understanding, data understanding, data preparation, modelling, and evaluation. Various combination of data preparation and modelling methods have been evaluated. The evaluation results show that the combination of feature selection and prediction model yields better results compared to prediction model without feature selection. The highest accuracy is achieved by Random Forrest at 97.82%, which is followed by Decision Tree at 97.06%, and Naive Bayes at 90.62%. This result indicates that a prediction model can be reliably used to predict customer churn in a telecommunication company. [ABSTRACT FROM AUTHOR]

Recency, Frequency, and Monetary (RFM) are behavior-based models used to predict customer behavior based on shopping activities. RFM model determines a customer's value for the company, which has been widely applied in various fields, especially the field of marketing. The RFM model was integrated with data mining to support customer segmentation and classification. The data mining techniques used were Self Organizing Maps (SOM), K-means, and Decision Tree C4.5. The framework was tested with 562 data of online customers who have purchased items in Indonesia's e-commerce platform. Three customer segments were distinguished, namely High Value Customer, Medium Value Customer, and Customer at Risk. The measure of goodness of segments formed was evaluated with the Davies-Bouldin index (0.472), cross validation (81%) and confusion matrix (74.48%). [ABSTRACT FROM AUTHOR]

The main objective of this research paper is to overcome the issue of imbalances in data and achieve an improved outcome for grouping by using synthetic minority over-sampling (SMOTE). The imbalanced issue with the dataset causes classification efficiency loss in many applications of data mining comprising of pattern detection, document categorization, and knowledge filtering activities. A sampling-based algorithm SMOTE is used herein to boost the efficiency of emotion classification that over-samples minority class instances to amount of majority class instances. YouTube© dataset was optimized using the SMOTE methodology and evaluated by the use of 3 machine learning algorithms, including decision tree (DT), multinomial Naïve Bayes (MNB), and support vector machines (SVM). As a consequence, SVM reaches full specificity with 92.20 percent filtering function and 88.33 percent filtering process classification. The oversampling solution, however, might not be successful because it merely replicates minority class instances without any additional details. A way to create a set of instances comparable to those in the minority class with high potential to identify with the minority class was proposed in order to address this issue, and use them instead in the oversampling phase. [ABSTRACT FROM AUTHOR]

The medical dataset is growing on a daily basis and there is a need to efficiently use this data to assist in the diagnosis of vast multitude of diseases. With the advancement in data mining, it is possible to extract interesting patterns from ever growing datasets in order to help take wise medical decisions. Diabetes is a chronic disease, caused due to either insufficient production of insulin by the pancreas or the cells of the body are not responding properly to the insulin produced. It leads to complications that affects heart, kidney, nerve and blood vessel damage, blindness and hence needs to be diagnosed at an early stage. This research work analyses different pre-processing tasks and identifies the best one that performs better than all the other techniques by constructing various classifier models. The classification model categories whether a person has diabetes or not. The four classification models explored were ANN, ID3, C4.5, and CART; compared the models based on evaluation parameters like Accuracy, Sensitivity, Specificity and Precision. [ABSTRACT FROM AUTHOR]

The Data mining algorithms have become vital to researchers in science, engineering, medicine, business, search and security domains. In recent years, there has been a terrific raise in the size of the data being collected and analyzed. Classification is the main difficulty faced in data mining. In a number of the solutions developed for this problem, most accepted one is Decision Tree Classification (DTC) that gives high precision while handling very large amount of data. This paper presents VLSI implementation of flexible architecture for Decision Tree classification in data mining using c4.5 algorithm. [ABSTRACT FROM AUTHOR]

Steel has found exhaustive engineering applications due to the abundant availability of its main constituent (Fe) in the earth‟s crust in the form of Fe2O3. It is produced with the requisite addition of carbon with iron. It can be made to exhibit a great variety of microstructures and thus has a wide range of mechanical properties to meet the demands of diverse engineering applications. Among enormous steel specifications, plain carbon steel accounts for more than 90% of the total steel production due to its favorable ductility, toughness and elongation properties, which also make it compliant to simple heat treatment operations. The objective of this paper is to investigate the effects of chemical composition (carbon, manganese, phosphorus, sulphur and iron) of plain carbon steel on its various mechanical properties using one of the data mining techniques, i.e. classification. The decision tree analysis is performed using classification and regression tree (C&RT) algorithm to identify the most significant chemical elements influencing machinability, Brinell hardness, tensile strength and elongation of plain carbon steels. The overall classification accuracy and prediction risk are estimated to explore the performance of C&RT algorithm. [ABSTRACT FROM AUTHOR]

Soil nutrient is the important factor to decide which crops is suitable to grow in that particular region. Data mining is used to solve various problems in agricultural sector. Soil fertile is identified by analyzing the availability of the soil nutrients level to grow suitable crop in that particular soil and also it is useful to find the deficiency of soil nutrients, finding the Soil Nutrients level will help the farmer to find the deficiency, based on the soil nutrients range farmer are recommended to cultivate the suitable crops in that particular soil. This research work focuses on J48 and C5.0 Decision Tree Algorithm to classify the soil data based on the soil nutrients and to predict the suitable crop for Shoolagiri Block at Krishnagiri District. [ABSTRACT FROM AUTHOR]

Tourism industry in Malaysia has been customarily thought and advanced towards universal markets since its early stages arrange in 1960s. Currently, study about tourism knowledge discovery is very little being addressed. The previous studies are still insufficient to extract important insights from tourism data within Malaysia context. Therefore, this paper aims to analyze profiles of tourists using data mining decision tree techniques where several combinations of the number of branches (2 and 3 branches) and different target splitting rules (Entropy, Gini, and Probability Chi-square) have been applied on comprehensive survey data and to find out the best performing algorithm among the six models for tourism knowledge discovery. Results show that there are a various type of tourists with each group having different patterns or rules. This research study can be very helpful for tourist association, hospitality and hotel managers. [ABSTRACT FROM AUTHOR]

Rainfall prediction is a challenging problem in the meteorological department around the world due to the accurateness of prediction. This paper studies on data mining techniques to predict rainfall using meteorological data of Subang Weather Station collected from January 2009 to December 2016. The data preparation process involves five weather factors which are maximum temperature, minimum temperature, evaporation, wind speed and cloud with 2922 observations. Predictive Decision Tree model, Artificial Neural Network model and Naïve Bayes model are developed for rainfall prediction and comparison. Surprisingly, results show that the performance of Decision Tree model is better as compared to the other predictive models with the misclassification rate of 0.15 and RMSE=0.35. Given enough set of data, rainfall can be predict using the data mining techniques. [ABSTRACT FROM AUTHOR]