Breast Cancer Patients - EDA and A/B Testing

Dataset: Kaggle

Objectives

Explore the dataset and extract insights from the data. Using statistical evidence to prove that

• Tumour stages are dependent of HER2-Status.
• There is a relationship between histology and HER2-Status.
• The effect of Proteins on patients' surival by tumour stages.
• Surgery performed on patient is dependent on his/her tumour stage
• The gender distribution is highly skewed
• Patients' survival curve differs by tumour stages

Conclusions

1. Tumour Stage-II's HER2 postive status has smaller proportion than Tumour Stage-III's HER2 postive status.

   P-value: 0.01

   

2. Surgery Procedures greatly depend on the Patients' Tumour Stages.

   P-value: 0.000022

   

3. Tumour Stages are dependent on different Age groups.

   P-value: 0.01

   

4. Although, we can't draw a definite conclusion that the male-patients/female-patients ratio is small due to smaller sample size (of male patients) but intuitively, the sheer difference speaks in it's favour.

   P-value: Nil | Since A/B Test is not applicable due to smaller sample size.