A research project aimed at predicting the psychological well-being of engineering students through machine learning techniques.
This study focuses on evaluating and forecasting the psychological well-being of engineering students by using various machine learning algorithms. It uses responses from a structured questionnaire designed to measure multiple psychological dimensions. The study aims to classify students as psychologically “well” or “unwell” and identify the most influential factors affecting their mental health.
- To assess psychological well-being using standardized survey data.
- To develop machine learning models to predict student mental health outcomes.
- To determine key features influencing psychological states.
- Data collected through a questionnaire administered to 202 engineering students.
- Questions cover areas like:
- Personal well-being
- Academic pressure
- Social and emotional conditions
- Physical and lifestyle habits
Note: The survey was conducted anonymously and ethically.
- Data Collection: Structured psychological questionnaire (based on standard scales).
- Preprocessing:
- Null value removal
- Normalization
- Label encoding
- Feature Selection: Pearson correlation and variance thresholding.
- Models Used:
- Decision Tree
- Logistic Regression
- K-Nearest Neighbors (KNN)
- Random Forest
- Support Vector Machine (SVM)
| Machine Learning Model | Accuracy |
|---|---|
| Decision Tree | 91.67% |
| Logistic Regression | 86.11% |
| K-Nearest Neighbors | 81.94% |
| Random Forest | 91.67% |
| Support Vector Machine | 88.89% |
- Best Performing Models: Decision Tree and Random Forest (both at 91.67% accuracy)
- Features such as emotional state, motivation, academic stress, interpersonal relations, and lifestyle patterns showed significant correlation with mental well-being.
- Models were able to predict student well-being status effectively with high accuracy.