Facilitates disease prediction by diagnosing user symptoms.
OVERVIEW
The multiple disease prediction system is a machine learning-based web application. It makes use of machine learning algorithms to analyze user symptoms and determine whether or not the user is suffering from a particular disease. Existing systems that facilitate early disease detection make use of either technical parameters or questionnaires. These existing formats are tedious or are presented in a way that is difficult for users to understand. Given the sensitivity of the issue at hand, namely disease prediction, this application has employed simple worded symptoms as a crucial construct.
The web-application has been developed using the Streamlit framework. Streamlit is a Python framework that makes it easy to create web applications for data science and machine learning projects. It facilitates the creation of interactive web apps with just a few lines of code, and it's designed to be intuitive and easy to use.
Installation procedure for Streamlit-
Streamlit can be installed using pip.
pip install streamlit
pip install streamlit-option-menu
The above commands help install streamlit and streamlit-option-menu. Streamlit requires Python 3.6 or a higher version.
Streamlit-option-menu facilitates the design of a multi-web page application and customizing the appearance of the sidebar.
WORKING SUMMARY
The multiple disease prediction provides a default 5 symptom entry display. It mandates that the user enter atleast 3 symptoms. Repetition of symptoms is not allowed. Additionally, the user can also enter more symptoms by requesting more input/select boxes.
Three machine learning algorithms have been utilized. They are: 1. Decision Tree 2. Random Forest 3. Naive Bayes
The symptoms entered are fed as input to each of the aforementioned algorithms. Each symptom select box has been provided with a default entry 'None'. The none input is not passed to the machine learning algorithms.
SOFTWARE USED
The Spyder IDE has been utilized to write the code for the multiple disease prediction system.
Installation of Spyder IDE: The preferred method of installation is to install Spyder using Anaconda. By default, Spyder is included in Anaconda.Anaconda can be installed from https://www.anaconda.com/. https://docs.anaconda.com/anaconda/install/windows/ provides detailed documentation regarding Anaconda installation for the windows platform.
LIBRARIES
The following libraries have been utilized:
1. Numpy : For passing symptoms to the machine learning algorithms.
2. Pandas: For accessing and working with the datasets.
3. Streamlit: To develop the web-application.
4. streamlit_option_menu: Used for designing the option-menu.
5. streamlit.components.v1: For rendering html and css code to make our web-app visually appealing.
6. cv2: To render the images on our web-application.
7. sklearn: Used to import machine learning algorithms for processing symptoms.
Installation-
pip install numpy
pip install pandas
pip install opencv-python
pip install scikit-learn
DEPLOYMENT
There are two ways of deploying the Streamlit app. It can be deployed on a local machine or can be made public. Deployment using localhost is achieved using the Streamlit run command.
streamlit run "C:\Users\Krishna\Downloads\DISEASE\website.py"
The above command helps deploy the Streamlit web-app using the localhost. This command can be executed using the Anaconda prompt.
Making the app public- Streamlit's Community Cloud and Streamlit Share provide a provision for deploying the app using Streamlit Cloud. The app is deployed on Streamlit Share. The Streamlit Share platform makes use of GitHub repositories. During runtime, the app fetches/pulls the necessary files in order to render the application. The platform instantaneously downloads the libraries that have been utilized in our application. The libraries along with their corresponding versions have been mentioned in the requirements.txt file.
The app is available publicly using the link: https://chiragradhakrishna43-7-multiple-disease-predict-symptoms-hkn9sw.streamlit.app/