Predictive methods for cancer based on machine learning

Principal Investigator

Name
Ashwini Suriyaprakash

Degrees
None

Institution
Independent

Position Title
Student

Email
ashwisp@gmail.com

About this CDAS Project

Study
PLCO (Learn more about this study)

Project ID
PLCO-292

Initial CDAS Request Approval
Aug 8, 2017

Title
Predictive methods for cancer based on machine learning

Summary
The project aims to analyze the screening data related to lung, colorectal and prostate cancers, and the digitized x-ray images from the lung screening trials, and determine the main risk factors that can predict each type of cancer using machine learning methods.

Aims

Using machine learning methods, this project will investigate each feature and their combinations in the screening data for each of the 3 types of cancer, and the characteristics of the digitized x-ray images from the lung screening trials. The aim is to determine the amount of training needed to predict each type of cancer along with the accuracy of prediction. I will also investigate if the data is enough for good prediction. Open source machine learning software is planned to be used with some customization.

Collaborators

None.