Comparative Examination of Decision Tree Classification Algorithms (KEY IJP************662)

• M Nancy

Abstract

The volume of data in educational databases is growing rapidly, holding hidden insights for improving student performance. Data classification, a key technique in data mining and knowledge management, groups similar data objects together. Among classification algorithms, decision trees are popular due to their simplicity. However, traditional algorithms like ID3, C4.5, and CART are limited to small datasets stored entirely in memory. This issue is overcome by SPRINT and SLIQ algorithms, which efficiently handle large databases. In our study, we compare these algorithmsperformance using existing datasets, with SPRINT showing the highest accuracy.