Student Adaptability Prediction in Online Education by Maria Saif Student Adaptability Prediction in Online Education by Maria Saif

Student Adaptability Prediction in Online Education

Portfolio project
Maria Saif

ML-random-forest-algorithm-

Student Adaptability Level Prediction (Online Education)

This project focuses on predicting the adaptability level of students in an online education environment using machine learning techniques. The objective is to classify students into one of three adaptability levels — Low, Moderate, or High — based on demographic, environmental, and technological factors.

Dataset

File: students_adaptability_level_online_education.csv
Records: 1205 rows
Features: 14 columns

Input Features:

Gender
Age
Education Level
Institution Type
IT Student
Location
Load-shedding
Financial Condition
Internet Type
Network Type
Class Duration
Self LMS
Device

Target Variable:

Adaptivity Level: Low, Moderate, High

Technologies Used

Python
pandas, numpy, matplotlib, seaborn (Data analysis & visualization)
scikit-learn (Machine learning & preprocessing)
joblib (Model serialization)

Preprocessing Steps

Encoding:
Ordinal Encoding: Age, Education Level, Class Duration
One-Hot Encoding: All other categorical features
Scaling:
StandardScaler applied to numerical data
Data Split:
80% Training, 20% Testing

Machine Learning Models

Several classification algorithms were tested:
Model Accuracy Logistic Regression 72.61% K-Nearest Neighbors 76.76% Support Vector Machine 82.16% Decision Tree 86.72% Random Forest 89.21% Gradient Boosting 83.40%

Hyperparameter Tuning

A GridSearchCV was performed on the Random Forest model to find the best hyperparameters.
Best Parameters:
{
  'bootstrap': False,
  'max_depth': None,
  'min_samples_leaf': 1,
  'min_samples_split': 2,
  'n_estimators': 300
}

0 likes · 6 views · Jul 21, 2025

Predicted student adaptability in online education using machine learning.

Tags