Breast Cancer Prediction

This is a machine learning project on the breast Cancer dataset to predict if a person is likely to be diagnosised with breast cancer or not. In this I have tested three very basic models to predict the diagnosis. The models are namely logistic regression, decision tree and random forest.

Libraries used

This project is entirely done on jupyter notebook in python. I have used the following python libraries :

1. pandas to read and visualize the data
2. seaborn to plot the dataset
3. matplotlib.pyplot to plot the graph
4. sklearn.preprocessing.StandardScaler for feature scaling
5. sklearn.linear_model.LogisticRegression for model selection

How to implement the project ?!

Steps to implement the project

1. Open google colab or jupyter notebook.
2. Either download the source code or copy the codes to a new .py file or new .ipynb file
3. run the codes

Accuracy of the model

I have tested it for three models here are the score for different models for both training and testing set.

Model	training Score	testing Score
Logistic regression	0.989	0.956
Decision Tree	1.0	0.938
Random Forest	0.997	0.973

As evident the decision tree model is overfitted and we get the best results with the random forest model. So finally Random forest should be used for future predictions.