Open AccessPerformance Evaluation of Regression Analysis Algorithms in Predicting the Enrollment of Basic Education
Author(s) -
Elizalde Lopez Piol,
AUTHOR_ID,
Luisito Lolong Lacatan,
Jaime P. Pulumbarit
Publication year - 2022
Publication title -
international journal emerging technology and advanced engineering
Language(s) - English
Resource type - Journals
ISSN - 2250-2459
DOI - 10.46338/ijetae0122_17
Subject(s) - linear regression , regression , regression analysis , statistics , mathematics , mean absolute error , mean squared error , computer science , econometrics
The use of Linear Regression in predicting enrolment has been shown to be beneficial, although it varies with various datasets and attributes; varying weights of the correlation of the attributes can be discarded if they do not impact the prediction. Data collecting had grown since prior investigations, resulting in a more complicated dataset with many varieties. As a result of the data being created by multiple clerks, cleaning and combining proved tough; nonetheless, the fundamental parameters remain intact. Different algorithms were examined but Linear Regression obtained the highest accuracy with a 12.398 percentage for the absolute error and a root mean squared of 26.936 to create a tangible model to anticipate the enrolment of Region IVA CALABARZON in the Philippines. This demonstrates that it was 2.067 percentage points more than the prior research.