Classifying heart diseases based on heart random numbers
From ISLAB/CAISR
| Title | Classifying heart diseases based on heart random numbers |
|---|---|
| Summary | Heart signal has entropy and can be used to generate random numbers. The idea is that, given a bunch of random numbers, we should predict if the source suffer from a desease |
| Keywords | Biometrics |
| TimeFrame | |
| References | |
| Prerequisites | Python, Matlab, ML |
| Author | |
| Supervisor | Pablo Picazo |
| Level | Master |
| Status | Finished |
It has been proven that the heart signal has some entropy and can be used to generate random numbers. These random numbers can be later used as part of cryptographic protocol such as secret schemes or authentication protocols.
In previous work, we demonstrated that random numbers derived from the heart dignal systematically fail some random tests (e.g., NIST). The idea of this project is to derive a bunch of random numbers from people with and without different heart diseases and train a machine learning algorithm that given a bunch of random numbers, it successfully predicts/classifies if they belong to a healthy patient or if not, which desease the patient is likely to suffer from.
