ANALISIS KETEPATAN WAKTU LULUS BERDASARKAN KARAKTERISTIK MAHASISWA FEM DAN FAPERTA MENGGUNAKAN METODE CHART

ANALISIS KETEPATAN WAKTU LULUS BERDASARKAN KARAKTERISTIK MAHASISWA FEM DAN FAPERTA MENGGUNAKAN METODE CHART
复制标题

分析 KETEPATAN WAKTU LULUS BERDASARKAN KARAKTERISTIK MAHASISWA FEM DAN FAPERTA MENGGUNAKAN 方法图表

DOI:
--
复制
发表时间:
2013
期刊:
影响因子:
--
通讯作者:
Yenni Angraini
Yenni Angraini
中科院分区:
--
文献类型:
--
作者:
Fira Nurahmah Al Aminy;Indahwati;Yenni Angraini

文献摘要

被引文献

相似文献

按时毕业是学生成功获得本科学位的指标之一。先前使用CHAID (Chi-Squared Automatic Interaction Detection,卡方自动交互检测)方法的研究得出,IPB学生准时毕业的最高概率分类来自FEM,最低概率分类来自FAPERTA。据此,本研究的目的是通过CART方法从准时毕业学生人数最多和最少的教师中确定学生的特征作为及时性通过的指标。FEM的分类树将准时毕业率最高的75%的学生的特征分类为IE系和MAN系的学生。FAPERTA的分类树将准时毕业率最高的26.1%的学生的特征分类为AGH和PTN系的学生,GPA大于3.335。FAPERTA分类树较好地估计了按时毕业的学生,准确率为81.47%;FEM分类树较好地估计了未按时毕业的学生,准确率为71.4%。关键词:chaid, CART,树形分类
Graduate on time is one of the indicators of a student’s success in obtaining their undergraduate degrees. Previous study using CHAID (Chi-Squared Automatic Interaction Detection) method concluded that the highest probability classification for the IPB students who graduated on time comes from FEM and the lowest probability classification comes from FAPERTA. According to that, the purpose of this research is to identify the student’s characteristics as an indicator of the timeliness pass from the faculty with highest and lowest number of students who graduated on time using CART method. Classification tree of FEM classifies the characteristics of the highest students that graduated on time by 75% as the students from Departements of IE and MAN. Classification tree of FAPERTA classifies the characteristics of the highest students that graduated on time by 26.1% as the students from Departments of AGH and PTN and have GPA greater than 3.335. The classification tree of FAPERTA is better to estimate students who graduated on time with accuracy value 81.47% and the classification tree of FEM is better to estimate students who didn’t graduate on time with accuracy 71.4%. Keywords-CHAID, CART, tree classification