Healthcare

Vaccine refrigerator (Internet of things)

This Vaccine Refrigerator solution is an educational project to showcase how an ordinary home refrigerator is transformed into a smart vaccine refrigerator that performs automatic temperature monitoring and recording, abnormal temperature alert, real-time inventory level monitoring, and alerts for vaccine low-stock level or impending stock expiration. This solution uses the Internet of Things (IoT) concept and is performed on the Amazon Web Services (AWS). It uses AWS lambda functions to simulate the mock temperature sensor and bar code scanner for inventory management. Other AWS resources such as AWS IoT, AWS Simple Notification Service (SNS), and AWS DynamoDB are also used. Finally, Python code on Jupyter Notebook is used to process the data collected into various visualisations of temperature recordings and vaccine stock inventory level.

All AWS and Python code & slides available at:

covid-19 health declaration (robotic process automation)

This Covid-19 Health Declaration solution is an educational project to help companies monitor the health status and work location of its employees. It uses UI Path’s Robotic Process Automation to manage the online declarations of the employees, process and compile the results for the company’s human resource administrator. This project uses UI Path’s functionalities on Chrome browser, data scrapping, Excel data manipulation, reading and saving files, and email.

All UI Path code, configuration files and slides are available at Github folder:

Cervical Cancer Classifier

This is a machine learning classifier of whether a patient has or does not have cervical cancer based on certain risk factors. The code is written in Python and executed on a Jupyter Notebook. The dataset is from the University of California Irvine (UCI) Machine Learning Repository, comprising 35 attributes of demographic information, habits, and historic medical records of 858 patients collected at Venezuela hospital. Different machine learning algorithms were used, and overall, Random Forest & Gradient Boosting are neck to neck with accuracy, precision & recall at 90%.

Python code & slides available at: