Indian Journal of Science and Technology
DOI: 10.17485/ijst/2016/v9iS1/109834
Year: 2016, Volume: 9, Issue: Special Issue 1, Pages: 1-4
Original Article
Ha-Ra Oh, Jung-Hyok Kwon and Eui-Jik Kim*
Department of Convergence Software, Hallym University, South Korea; [email protected]
jhkwon@hallym. ac.kr
[email protected]
*Author for correspondence
Eui-Jik Kim
Department of Convergence Software
Email:[email protected]
Objectives: This paper presents an implementation of a menstrual irregularity prediction model based on a decision tree for a big data healthcare system. Methods: To build the menstrual irregularity prediction model, we use personal health data classifying people into the menstrual irregularity or normal groups as the training dataset. For accurate prediction, we use various attributes that affect menstrual irregularity, such as age, irregular meals, childish diseases, and accidents or serious trauma. For data classification, we create a decision tree by selecting the most influential attribute in each decision phase. The modeling and performance evaluation are performed through R Studio Version 3.2.4. Findings: We evaluate the performance of the menstrual irregularity prediction model through a confusion matrix. The evaluation results show that the menstrual irregularity prediction model exhibits 84.0 %, 82.9 %, and 85.5 % of accuracy, precision, and recall performance, respectively. Improvements/Applications: We expect that our menstrual irregularity prediction model will be a reference guideline for realizing the big data healthcare system.
Keywords: Big Data, Classification, Decision Tree, Healthcare System, Menstrual Irregularity Prediction
Subscribe now for latest articles and news.