dropout-prediction

My Own Repository with workfolders and others such as Group Projects etc...

Machine Learning project to predict student dropout using the OULAD dataset and Random Forest/KNN models

Predicts whether an undergraduate student will drop out or graduate using logistic regression and decision tree models on Portuguese higher education data. Log transformations are kept only where they reduce skewness. Logistic regression reaches 0.930 accuracy and 0.967 ROC-AUC against a 0.609 baseline.

PyTorch and TensorFlow implementations of LSTM, Transformers, and autoencoders for modeling sequential student behavior.

End-to-end supervised ML pipeline for student performance and dropout prediction. Features data preprocessing, feature engineering, multi-model training, performance comparison, model persistence, and reproducible project structure with detailed README files.

Predicting student dropout using XGBoost and Neural Network models on online course data

Machine learning system achieving 87% accuracy in predicting student dropout risk using Python and scikit-learn

Detecção Precoce de Estudantes em Risco de Evasão Usando Dados Administrativos e Aprendizagem de Máquina

Generative, semi-supervised and fair learning for student-dropout prediction — a reproducible, multi-seed, significance-tested benchmark on UCI Realinho + OULAD. Code/results supplement to the technical report.

Turning an accepted research paper into a working AI product: early dropout prediction for Learning Management Systems, with per-student explanations, calibrated uncertainty and fairness measurement. Synthetic data, decision support only.

AI Based Dropout Prediction and Counselling System using Python , Flask and Machine learning.

Agent-based simulation environment for Open & Distance Learning (ODL) research.

Proyecto Final de Almacenes y Minería de Datos - Predicción de deserción con Random Forest.

Proyek Akhir Belajar Penerapan Data Science Dicoding: Sistem peringatan dini (Early Warning System) berbasis Machine Learning untuk memprediksi risiko dropout mahasiswa Jaya Jaya Institut menggunakan Streamlit.

An explainable machine learning system for student dropout risk prediction using Streamlit, ensemble models, and SHAP.

AI-based web application that predicts students at risk of dropping out and provides personalized counseling recommendations based on academic and student-related factors.

Framework para benchmarking reprodutível de arquiteturas de dados (DuckDB, Dask, Polars) com verificação de anti-leakage temporal. Demonstrado com predição de evasão escolar municipal (INEP & WorldBank, 2007-2024).

Comparative study of deep learning vs gradient boosting for student dropout prediction on the OULAD dataset, with SHAP interpretability and a three-tier early warning system.

End-to-end data science project combining supervised classification (Random Forest), cohort clustering, and local RAG via Ollama to predict and mitigate student dropout risks.