Disease Classifier

A deep learning-based project designed to classify diseases based on symptoms. This repository includes scripts for preprocessing datasets, creating optimal data representations, and training a deep neural network model to predict diseases.

Table of Contents

• Overview
• Dataset
• Features
• Requirements
• Installation
• Usage
• Files and Scripts
• Model
• Future Work
• Contributing
• License

Overview

Disease Classifier is a tool designed to assist in identifying potential diseases based on a user’s symptoms. The system leverages a deep neural network trained on preprocessed symptom-disease data to provide predictions. This project is intended for educational and exploratory purposes.

Dataset

The project includes the following datasets:

1. new_df.csv: A refined dataset prepared for machine learning.
2. dataset.csv: The original dataset containing disease-symptom relationships.
3. final_df.csv: The processed dataset used for training the classifier.
4. Symptom-severity.csv: Details the severity of each symptom.
5. symptom_Description.csv: Contains descriptions of symptoms.
6. symptom_precaution.csv: Lists precautions to take for each disease.

The creating an optimal dataset.ipynb script processes these datasets to produce final_df.csv, an optimized representation of the data used for training the classifier.

Features

• Preprocesses and cleans raw datasets by standardizing symptoms and removing inconsistencies.
• Construct a deep learning model using TensorFlow and Keras to classify diseases.
• Output predictions based on symptom input, with the top prediction highlighted.

Requirements

The project requires the following Python libraries:

• pandas
• numpy
• scikit-learn
• tensorflow
• keras
• joblib

Installation

1. Clone this repository:

   git clone https://github.com/your-username/disease-classifier.git

2. Navigate to the project directory:

   cd disease-classifier

3. Install dependencies manually using pip:

   pip install pandas numpy scikit-learn tensorflow keras joblib

Usage

Step 1: Preprocess Data

Run the creating an optimal dataset.ipynb script to preprocess and generate the final_df.csv dataset:

• Cleans the original dataset by standardizing and encoding symptom names.
• Saves the processed dataset for training the model.

Step 2: Train the Model

Execute the disease_clf_deepLearning.ipynb notebook to:

• Load the preprocessed dataset.
• Train a deep learning model on symptom-disease data.
• Save the trained model for later use (optional).

Step 3: Predict Diseases

Modify the prediction block in disease_clf_deepLearning.ipynb to test new symptom inputs. The model outputs the most probable disease based on the given symptoms.

Files and Scripts

• creating an optimal dataset.ipynb: Preprocesses datasets to create final_df.csv.
• disease_clf_deepLearning.ipynb: Builds and trains a deep learning model, and makes predictions.
• Symptom-severity.csv: Provides information on the severity of each symptom.
• dataset.csv: The original dataset containing disease-symptom relationships.
• final_df.csv: The processed dataset used for training the classifier.
• symptom_Description.csv: Contains descriptions of symptoms.
• symptom_precaution.csv: Lists precautions for each disease.

Model

The neural network model consists of:

• Input Layer: Accepts 131 symptom features.
• Hidden Layers:
  • Dense layer with 132 nodes and sigmoid activation.
  • Dense layer with 50 nodes and ReLU activation.
  • Dense layer with 17 nodes and ReLU activation.
• Output Layer: Dense layer with 41 nodes and sigmoid activation, representing the number of diseases.

The model is compiled with:

• Optimizer: Adam
• Loss Function: Binary cross-entropy
• Evaluation Metric: Accuracy

Future Work

• Improve model accuracy by experimenting with hyperparameters and architectures.
• Add support for multi-class disease predictions.
• Develop a user-friendly interface (e.g., web or mobile app) for interacting with the classifier.

Contributing

Contributions are welcome! Please open an issue or submit a pull request for any bugs, improvements, or feature requests.