Data mining and predictive analysis on the employment of fresh graduate students in public universities in Malaysia
This study was aimed to use data mining to predict the employment of fresh graduate students in public universities in Malaysia. The research design of the study was model development using Rapid Miner Studio. The model used both supervised and unsupervised machine learning algorithms including k-Ne...
Saved in:
| Main Author: | |
|---|---|
| Format: | thesis |
| Language: | eng |
| Published: |
2020
|
| Subjects: | |
| Online Access: | https://ir.upsi.edu.my/detailsg.php?det=6915 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study was aimed to use data mining to predict the employment of fresh graduate students in public universities in Malaysia. The research design of the study was model development using Rapid Miner Studio. The model used both supervised and unsupervised machine learning algorithms including k-Nearest Neighbor (kNN), Nave Bayes, Decision Tree, Logistic Regression, Support Vector Machine (SVM) and Neural Network. The sample consisted of 16,729 fresh graduate students were collected from the Tracer Study Unit of Ministry of Higher Education (MOHE). In order to build the classification model, Cross Industry Standard Process for Data Mining (CRISP-DM) methodology was applied. For the evaluation, 70% of the dataset were used as the training set and the remaining 30% were used as a testing set. To determine the error rate and to justify the accuracy of the proposed model objectively, classification error was used as the evaluation metric. The key finding of the predictive analysis revealed that employability among fresh graduate students can be predicted with 59.90% accuracy with a Neural Network as the most accurate predictive model. The significant factors contributing to graduates employment were problem-solving and decision making skills. The unemployment, on the other hand,vwas mainly attributed to these factors poor English competency, majoring in Malay, Education, and Science fields. In conclusion, the empirical data supported Neural Network model for predicting the employability among fresh graduate students in which the graduates should possess critical skills such as problem-solving and decision making skills. In implication, the predictive model was useful for graduate students, management of public institutions, Ministry of Higher Education, human resource personnel and academic staff in predicting the graduates employability. |
|---|