International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395 -0056
Volume: 04 Issue: 03 | Mar -2017
p-ISSN: 2395-0072
www.irjet.net
ASSESSMENT OF DECISION TREE ALGORITHMS ON STUDENT’S RECITAL Arundthathi A1, Ms. K. Glory Vijayaselvi2, Dr. V. Savithri 3 1
Post graduate Student, M.Sc. CST Women’s Christian College, Tamil Nadu, India 2 Assistant Professor, M.Sc. CST, Women’s Christian College, Tamil Nadu, India 3 Assistant Professor, M.Sc. CST, Women’s Christian College, Tamil Nadu, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Data mining is a dominant progression to
Decision tree is a renowned classification technique commonly used in many researches. Decision tree model represents a flowchart-like structure where each node represents a test on data objects and the leaf node represents the class label. Decision tree are used in all domains to predict hidden patterns. Decision tree is wellknown because it is simple and easy to interpret.
forecast upcoming behaviors. Data Mining is used in various domains and disciplines to solve an existing problem or to envisage comportments. The different Data mining techniques are Clustering, Association, Classification, Regression and Structured Prediction. Classification is a most important task of simplifying dataset. Decision tree method is commonly used in Classification technique. Decision tree model is represented by branch and nodes. There are several Decision tree algorithms to contrivance in data mining tools. The objective of this study is to compare the most frequently used Decision tree algorithms in various domains and holds the good in predicting the best decision tree algorithm. Educational dataset is implemented to find the accuracy of the Decision tree algorithms and to predict the student’s performance level. This research provides an idea to educators on student progress level.
The aim of this research is to compare the efficiency of different decision tree algorithms. Education is one of the domains which is profited by Data mining. To compare the decision tree algorithms Educational dataset from a reputed college is implemented. Semester marks of college students is collected and analyzed by Data mining tool to classify students to Grade A, Grade B or Grade C and predict the next semester percentage of each student.
Key Words: Classification, Decision Stump, Decision
This study finds out the commonly used decision tree algorithms in various domains. The objective of this study is to list out the efficiency and accuracy of decision tree algorithms. College student’s performance in exam is analyzed to rank them and predict their future performance. This research helps professors to predict achievement levels and identify a student or a group of students in need of further attention.
Tree algorithms, J48, Hoeffding Tree, Random Forest, Random Tree, REPTree.
1. INTRODUCTION Data mining is a prevailing technology with prodigious prospective to emphasis the most essential information in data warehouses. Data mining tools are used to foretell future tendencies, making practical and knowledge-driven verdicts. Data mining tools can response queries that usually were complex to resolve. Data mining tasks is used to discover hidden patterns from existing information. Data mining is applicable for any kind of data repository. Data mining methods can be applied on existing software and hardware platforms to improve the significance of present information resources.
2. REVIEW OF LITERATURE Data mining is used in many researches for various purposes in different field. Researchers have worked in educational field to predict loyal students and dropout students to improve educational quality. Medical field is widely used with data mining to diagnosis many diseases like Breast cancer, Diabetics, Typhoid. In organizational field data mining is supportive to make decision and set marketing goal. In weather domain data mining commonly used to predict weather. In environment domain data mining is implemented and analyzed with soil, iris flower and mushroom datasets.
Classification is a data mining technique used to organize data objects according to given class labels. Classification process slants using training set of data in which all data objects are already classified with class labels. The classification algorithm absorbs the training set and constructs a model. The constructed model is used to classify unclassified large datasets. There are many classification algorithms.
© 2017, IRJET
|
Impact Factor value: 5.181
Nilima Patil, Rekha Lathi, and Vidya Chitre [1] provided the way for decision making process of customers to recommend the membership card using classification
|
ISO 9001:2008 Certified Journal
| Page 2342