Cancer Survival Prediction

For Chulalongkorn AI Academy candidate screening

• Compatible with Python 3.6 and 3.7

Problem introduction

• The ability to predict survival chance for cancer patient provides crucial information for making medical decision.
• Survival chance depends on patient and tumor characteristics.
• Patient basic characteristics were graded by clinicians through interviews and blood tests.
• Tumor characteristics were collected via imaging followed by manual annotations by doctors and automatic image processing.
• Can we predict the patient's survival chance with these information?

Getting Started

Run the following command in your Terminal to install Project Dependencies

python install -r requirements.txt

To begin training the project, download and rename the dataset from the Kaggle page to the data/folder as so:

data/radiomics_v1.csv # Clinical Data from 197 Patients
data/radiomics_v2.csv # 30 Patient with 3 Doctors Data

Create a Cancer Survival Estimator using the following command:

python cancer.py

Code Walkthrough

You can follow along the attached Jupyter Notebook titled cancer_survival.ipynb or compiled web-format cancer_survival.html on this project and see how data was treated before a Classification Model was trained on the data. Rule #1 of Data Science is Garbage In, Garbage Out. This means that we need to identify how to work with the data given first. The saved pickle file, cancer_survival_estimator.pkl can be used to predict chances of survival with great accuracy.

Sample Output:

[INFO] Number of Wrong Predictions: 7 / 22
[INFO] R2: -0.4667
[INFO] Mean Squared Error: 0.3182
[INFO] Root Mean Squared Error: 0.5641