Prediction of secondary public school selection based on parents’ preferences using naïve bayes / Mohamad Hafizi Masdar
Choosing a school for the secondary students is not a simple task nowadays since there are many factors needed to be considered by parents. These might be an overhead to the parents to filter each school and try to sort it out based on their requirements. Therefore, a predictive model which is a qua...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2017
|
| Online Access: | https://ir.uitm.edu.my/id/eprint/18157/2/TD_MOHAMAD%20HAFIZI%20MASDAR%20CS%2017_5.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Choosing a school for the secondary students is not a simple task nowadays since there are many factors needed to be considered by parents. These might be an overhead to the parents to filter each school and try to sort it out based on their requirements. Therefore, a predictive model which is a quantitative research that implemented artificial intelligence (AI) strategy using Naïve Bayes (NB) technique been applied in order to predict the chosen secondary school in Gombak based on parents’ preferences. Through six different independent variables (IV) namely distance of home from school (DSH), sibling enrolment in the same school (SE), highest parent’s level of education (PLE), household income (HI), employment status (ES) and race, 64 prediction models were produced and tested. Among all these models, the highest accuracy detected was the combination of four IVs which are DSH, PLE, ES and race, that resulted more than 55% of accuracy. This finding then implemented into a system which is Prediction of Secondary Public School System (PSPSS) to help parents making school selection easier since parents’ concerns are taken into consideration. NB provided a new way of solving a problem in determining the vital factors for making a school choice. Since this is the preliminary and exploratory study, the highlighted achievement is not the accuracy percentage of the prediction, but a new angle on the ability of how AI can solve the problem within the domain of this study. This finding can be further improved by using the same technique which is NB through modification of IV used and the number of data collected. Other AI techniques such as support vector machine (SVM), multiple linear regression (MLR) and artificial neural network (ANN) also can be used to compare with the current NB as well as an enhancement to readily available findings. This study also may be broadened to other regions or states to help more parents making school selection easier. |
|---|