Development of a Web-application(dashboard) for ML model inference and data analysis

Description

This project is an interactive web page to visualize and predict whether a person has suspected diabetes based on 22 signs. It includes several trained machine learning models and has a user-friendly interface. The web page is developed using Streamlit library and is hosted on streamlit.Cloud hosting platform

Installing

Through Streamlit.Cloud:

Visit the website: https://mlrgrkotov.streamlit.app

On your PC locally

1. Clone the Git repository
2. Restore the dependancies using pip install -r requirements.txt
3. Run web application using streamlit run hub.py

Functions

1. Validate the performance of machine learning models and metrics by loading your own dataset.
2. Prediction of diagnosis based on user-defined indicators.

Credits

Created by Kotov Sergey Vladimirovich, a student of OmSTU, Faculty of FIT&CS, Department of PM&FI, direction of study "Fundamental Informatics and Information Technologies", student group FIT-222 as a report on calculation-and-graphic work on the discipline "Machine Learning and Big Data". Year of 2024.