PALO IT Project: Wait No More, just DIETBEATIT!

Inspiration

My interest in this project is because it reminds me of past experiences with my grandparents. Both my grandparents were victims to the cause of late hospital admission and treatment which leads to deteriorating health issues. Additionally, my grandmother was a diabetic patient. I've seen the seriousness of diabetes could be when I witnessed my grandmother losing her leg because of diabetes. Hence, I've driven my passion to the second problem statement. I hope that my created app could devote some effect on raising public awareness to take serious attention to diabetic prevention.

What it does

The purpose of the project is to understand the root cause of the issue of late hospital treatment for patients, together proposing a supportive app for not just providing convenience to people to assess the GP clinic services and also raising public awareness on preventing chronic diseases like diabetes by providing related benefit plans, health tips, and related information.

How we built it

In my first solution, I performed EDA analysis on the hospital admission dataset and developed different machine learning methods for accurate predictions on patients' hospital staying durations. For the second problem statement, I've used Power App platform to create a health app named "Die.Beat.IT". Inside this app, I've equipped it with 4 domain features, including appointment booking with doctors, General Care benefit plan, browsing on GP clinics around Singapore, and Diabetes challenges (step-by-step challenge guide to prevent diabetes). The purpose of the project is to understand the root cause of the issue of late hospital treatment for patients, together in proposing a supportive app for not just providing convenience to people to assess the GP clinic services and also raising public awareness on preventing chronic diseases like diabetes by providing related benefit plans, health tips and related information.

Challenges we ran into

When solving the first problem statement in this project, due to some quota constraints for several attempts at Azure ML, I've decided to proceed with my first problem statement from scratch by using MS visual studio code. Secondly, I've some bug issue in my created Die.Beat.IT app. I'm still solving it and enhance my app performance by coming up with better fit features that could better fit my app functionality.

Accomplishments that we're proud of

In this project, I've successfully trained some models for predicting the hospital stay durations and detecting late treatment outcomes with good model performance and accuracy scores. I've believed models would help the healthcare partners in mitigating the hospital over-occupancies issue by having an accurate prediction for better resource allocations (hospital bed and staying time). Also, my analysis of the patient group segment would also help the healthcare companies to have more concerns about those potential groups of people that fail to receive prompt treatment when encountering emergency situations that escalate their health risk and a proper admission arrangement. Aside from that, I've believed that my development of the DietBeatIT app, which would benefit the community in assessing healthier lifestyles by regularly monitoring their health status and calorie intake. In addition, the DietBeatIT app has also provided convenient assess for the community to several services including providing well-organized information on the GP panel clinics nearby their areas, a more step-by-step focus guide for the user to avoid diabetes loopholes via Fight Diabetes Challenges and not to forget the online appointment with doctors features via the app.

What we learned

Even though I've had no experience in app development, but thanks to the Power App, i can build up my first ever health monitor app with low code requirements. Also, my analytics and coding skills have much improved by taking up the hackathon challenge. I've learned variety of approaches to data cleaning, analysis and modeling process in this project.

What's next for Wait No More, just Die.Beat.IT !

In the proceeding action, I've worked on improving the scalability and availability of the application to benefit public users. Also, I will improve my cloud skills and data pipeline skills to enable faster, more efficient, and more scalable performance of data cleaning, transformation, engineering, and model training tasks, which could escalate productive and effective performance results for my prospective employer and company.

Built With

• app-development
• data-analytics
• jupyter
• machine-learning
• ms-code-studio
• notebook
• powerapp
• python