Fuzzy C-means於學習風格分類進行課堂分組之研究

Categorizing Learning Styles for class project groups by Fuzzy C-means

陳品叡(Pin-Jue Chen)

指導教授 ： 楊康宏

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2021年)

https://doi.org/10.6840/cycu202100625

摘要

在過往學者的研究中發現，課堂分組的教學方式能在不同的適用情況下幫助學生提升學習能力，分組模式的不同也將有可能影響學習者在學習中的成效，同質性分組即為常見的分組方式，在過去，教學者多利用經驗或是簡單規則的方式對學生進行分組教學，但對於同質組或是異質組的差異及對應不同學習方式及程度的同學可能會造成學習效果的明顯不同，因此，本研究利用Fuzzy-C-means模糊分群方法，通過學習風格量表(Index of Learning Style )做為課前分組數據，提供教學者一個課前同質性分組的參考依據。本研究將藉由大量模擬常態數據，控制不同切分點、分組數、組人數、構面數量等參數，驗證演算法於不同情境的有效性，並建立平均隸屬度與分群錯誤率的關係，應用於真實課堂資料。結果顯示，分組數提升及組人數下降將導致分組錯誤率上升，構面數量從模擬結果則是不明顯影響錯誤率，在真實資料應用上，本研究使用簡單回歸方法找出平均隸屬度與錯誤率關係，並提供教學者一個確切分群效果的參考數值。

關鍵字

同質性分組 ； 模糊C均值 ； K-平均 ； 學習風格 ； 群集分析

並列摘要

In the research of past scholars, it was found that the teaching method of class grouping can help students improve their learning ability in different applicable situations, and the difference of grouping methods may also affect the effectiveness of students in learning. Homogeneous grouping is a common method of grouping. In the past, teachers used to group students with experience or simple rules in class, but the differences between grouping in homogeneous or heterogeneous with different learning methods or levels may cause significant differences in learning effects. Therefore, this study uses the Fuzzy-C-means clustering method grouping pre-class by index of learning style, offering teachers with a reference basis for homogeneity grouping. This research will use a large amount of simulated normal data to control parameters such as different segmentation points, number of groups, group size, number of dimensions, etc. Verifying the effectiveness of the algorithm in different situations and to establish the relationship between the average membership degree and the error rate of grouping, finally, applying in the real-class. The results show that the increase in the number of groups and the decrease in the group size will lead to the increase of the error rate of grouping, while the number of dimensions does not significantly affect the error rate from the simulation results. In the application of real data, this study uses a simple regression analysis to find the relationship between average membership degree and the error rate of grouping, and provide the teacher with a reference value of the clustering effect.

並列關鍵字

Homogeneous grouping ； Fuzzy-C-means ； K-means ； Learning style ； Clustering analysis

參考文獻

英文資料

Alfonseca, E., Carro, R. M., Martín, E., Ortigosa, A., Paredes, P. (2006). The impact of learning styles on student grouping for collaborative learning: a case study. User Modeling and User-Adapted Interaction, 16(3-4), 377-401.

Bezdek, J. C., Ehrlich, R., Full, W. (1984). FCM: The fuzzy c-means clustering algorithm. Computers Geosciences, 10(2-3), 191-203.

Blumenfeld, P., Soloway, E., Marx, R., Krajcik, J., Guzdial, M., Palincsar, A. (1991). Motivating Project-Based Learning: Sustaining the Doing, Supporting the Learning. Educational Psychologist, 26, 369-398.

Cerezo, R., Sánchez-Santillán, M., Paule-Ruiz, M. P., Núñez, J. C. (2016). Students' LMS interaction patterns and their relationship with achievement: A case study in higher education. Computers Education, 96, 42-54.