Using Casual Inference

In this project I am going to :

Use casual inference using Pearl's framework.
Infer the causal graph from observational data and then validate the graph;
Merge machine learning with causal inference;

Table of Content

Introduction
Data Exploration
Feature Importance

Introduction

Judea Pearl and his research group have developed in the last decades a solid theoretical framework to deal with that, but the first steps toward merging it with mainstream machine learning are just beginning.

Data Exploration

The features of the dataset were grouped into 3 features.The 'mean features','worst error features' and 'squared error features'. The correlation of how each variable in the group of features affected the 'diagnosis' was calculated and a bar plot showing feature correlations were plotted.

The outliers for each variable were identified and removed.

Feature Importance

Random Forest Classiffier was used to identify relative importance of the features to the 'diagnosis'. A random forest classfier was built and trained on the dataset. The Accuracy score was calculated.It achieved an accuracy score of 95%.This score was good enough for our model.'fetaure_imprtances_' was used to get the order of feature importance.

Causal Model