First Machine Learning Project

First Machine Learning Project

I have recently created my first Github repository bijay-05/end-to-end-ML-with-streamlit for sharing my first end-to-end-machine-learning project. Although I did some end-to-end-ml projects following video tutorials, those projects were not something that I did completely myself. But this project is the one which I did myself from beginning to the end. So, I thought a small blog describing about My First Machine Learning Project will make this first project more special in a sense, whenever I visit here I could view my first project and who knows, some inspiration might come for keep learning and doing projects.

The main idea behind this project is to build Diabetes detection system in interactive web application using streamlit. All the technical stuff about the project is available in the repository. Apart from technical stuff, this blog is about a high level overview of the project. Various exploratory data analysis and visualizations are performed on the data. Few machine learning models are trained after some feature engineering so that our model can train fast. Among them, svm (support vector machine) model performed the best. So we dumped the model to .pkl file and for deploying our web application, app.py file was created. To get inference on the input data from our web application, we loaded the model from .pkl file using pickle. Since streamlit hosts web application on local host, the last cell of the notebook includes code for localtunnel which helps to expose the local host and share our web application with others. I think this blog is not so good but I just write this blog to accompany the my first machine learning project and get engaged writing for sometime. Thank You!

The demo of the interactive web application is below. Screenshot (24).png