This project focuses on predicting depression using various machine learning algorithms. It explores the effectiveness of different models and techniques for analyzing healthcare data to identify individuals at risk of depression.
This project aims to develop and evaluate machine learning models for depression prediction, leveraging a dataset containing demographic, lifestyle, and medical history information. The primary goals include:
-
Data Exploration and Preprocessing: Understanding the dataset characteristics, handling missing values, and preparing data for model training.
-
Model Selection and Training: Evaluating different algorithms (e.g., Logistic Regression, Decision Trees, Random Forests, XGBoost) and optimizing their hyperparameters.
-
Model Evaluation and Comparison: Assessing the performance of each model using metrics like accuracy, precision, recall, F1-score, and ROC-AUC.
-
Feature Engineering and Selection: Investigating feature engineering techniques to enhance model accuracy.
-
Addressing Class Imbalance: Implementing strategies to mitigate the issue of class imbalance in the dataset.
-
Deployment and Visualization: Exploring potential deployment methods and creating insightful visualizations of the results.
This project utilizes a depression dataset and several Python libraries, including:
-
Dataset: Depression data with features such as age, marital status, education, income, smoking habits, physical activity, and medical history.
-
Libraries: Pandas, NumPy, Matplotlib, Seaborn, Scikit-learn, XGBoost, LightGBM, Imblearn, Optuna.
The project is structured as follows:
-
Code: Contains the Python code implementing data loading, preprocessing, model training, evaluation, and visualization.
-
Data: Stores the depression dataset used for the project.
-
README.md: Provides a detailed explanation of the project, its goals, and methods.
The trained machine learning model has been deployed for accessibility. You can access the deployment through:
-
Deployment - https://depa-depression-prediction-app.onrender.com
-
Deployment Video - https://youtu.be/xPEIinKdC9s
A detailed presentation about the project, including insights, findings, and deployment details, is available at:
- Presentation Video - https://youtu.be/JRXlteHIMcI
This project demonstrates the application of machine learning for depression prediction, highlighting the potential of data-driven approaches in improving mental healthcare. The project findings and insights can contribute to developing personalized treatment plans, early interventions, and better risk assessment strategies for individuals at risk of depression.
This Project has been built by - Dileep Naidu, Sreekar Navarasala, Praneeth Goud
Contributions to this project are welcome! Feel free to fork the repository, explore the code, and suggest improvements.
For any questions or collaboration opportunities, please reach out via dile2107@gmail.com.