Description

Machine learning project using logistic regression, written from scratch (without any librairy like scikit-learn). You've got two versions :

• One on jupyter notebook : cancer_prediction.ipynb
• One using streamlit, which is a web interface : cancer_web.py

Installation

In order to launch this project you'll need to install python, go to this official link https://www.python.org/downloads/.

• For the first version You can easily follow this link which explains very well how to install Jupyter Notebook, with the terminal : https://medium.com/analytics-vidhya/how-to-install-jupyter-notebook-using-pip-e597b5038bb1
• For the second Use the package manager pip to install the different packages.

  pip install streamlit
  pip install pandas
  pip install numpy
  pip install sklearn

Usage

Before starting, you can notice that the processing part is quite short, the reason being that I focused a lot more on the algorithm part knowing that the dataset was already prepared.

• For the first version You'll need to import both "cancer_prediction.ipynb" and "survey lung cancer.csv" into a document. After that, just launch Jupyter Notebook, and then open the cancer.ipynb file !
  The file is separated into two parts :
  • The processing part which consists in analyzing the data
  • The Logistic regression part where you'll find the code for this algorithm, from scratch (i hope that you've some math knowledge :))
• For the second Import "cancer_web.py" and "survey lung cancer.csv" into a document, enter in it, and launch the application with the following command:

  streamlit run cancer_web.py

Screenshots