datalordstephen/mock-system

This code contains code for a mock anomaly detection system in a hostpital setting. I wrote this code during my IT at RAIN for my 200lvl SIWES program.

Mock Anomal Detection System

This code contains code for a mock anomaly detection system in a hostpital setting. I wrote this code during my IT at RAIN for my 200lvl SIWES program.

This code was written to mimic an ideal anomaly detection system in a hospital system. The goal was to use this web app to present the idea of an anomaly system to other staff and explain the workings.

An ideal anomaly detection system would follow a machine learning approach, Clustering to be specific.

Background

In a hostpital, there is a system that allows doctors to view files of patients, prescribe medications and all of that from their computers. Doctors have different specializations, and these specializations are useful in different scenarios, depending on the patient's condition.

It would be weird if a doctor specializing in dentistry was to try to access the files of a patient suffering from asthma right? An anomaly detection system would be able to flag such an interaction, and record it in a database, to either refer back to in the future, or to use as proof to back up a suspiscion.

Usage

To run this code, python has to be installed.

• Install requirements

pip install -r requirements.txt

• Generate an .env file with a random secret key and database URI

py create_env.py

• Create a mock database with 20 patients and 1 doctor

py populate_db.py

• Run the app

py run.py

After running the app, 20 patients would appear on the screen, and clicking on any patient other that FLU patients would trigger a popup on the screen saying Suspiscious activity detected. The interaction would also be logged to the database, indicating the doctor, patient, and that the interaction is suspiscious.