Title:
A Hybrid Recurrent Networks with Semi-Supervised Support Vector Machines for Dropout Prediction and Personalized Intervention

Author:

Supervisor:

Department:

Computing

Year:

2026

URL:

Abstract:

Student dropout remains a persistent challenge in higher education, with consequences that extend beyond academic loss to financial and societal impact. Although predictive analytics has been widely a...
View more
Abstract Image

Multimedia Content:

01

Methodology

1. Variational Autoencoder (VAE) Data Augmentation To address the severe class imbalance commonly observed in student dropout datasets, the study developed a Variational Autoencoder-based augmentat...
View more
Methodology Image
02

Citation Analysis

1. Journal of Educational Data Mining (JEDM) Scope: Educational data mining, educational analytics, student modelling, and predictive learning systems. Relevance: Highly relevant because the thesi...
View more
03

Impact

1. Technical Impact This research introduces DeepS3VM, a novel hybrid framework that combines Recurrent Neural Networks (RNN) and Semi-Supervised Support Vector Machines (S3VM) for student dropout ...
View more
04

Benefits to Society

1. Environmental Benefits Although the research is primarily focused on education, increased adoption of intelligent digital learning support systems can reduce dependence on paper-based monitoring...
View more
05

Applications

1. Student Early Warning System Benefit: Identifies students at risk of dropping out before academic failure occurs. Implementation: DeepS3VM can be integrated into institutional analytics platfor...
View more
06

Infographic

Summary of this Thesis
Infographic Image
07

Patents / Products

Product Name: DeepS3VM: Hybrid RNN–Semi-Supervised SVM Framework for Student Dropout Prediction and Personalized Intervention Significance: DeepS3VM is the primary innovation of the thesis and pr...
View more
08

Experts

1. Ts. Dr. Aliza Bt Sarlan Institution: Universiti Teknologi PETRONAS (UTP), Malaysia Contribution: Main supervisor responsible for the development of the DeepS3VM framework and educational analyt...
View more