SharkHack 2020
These will work in anaconda prompt. Please follow the corresponding instructions for linux.
- Clone the respository using `git clone https://github.com/Tannybuoy/HeartsofSharks
- Redirect to that repository using
cd/Documents/GitHub/HeartsofSharks, or whatever the route is in your Desktop - Type
python "app.py"to open the website in your browser
Due to the fear of hospitals caused because of growing COVID19 cases, people have stopped visiting hospitals. This has led to reduction in detection and prognosis of diseases which require expertise of doctors, like cancers, heart diseases and more. It is the need of the hour for people to be able to detect whether they have a disease or not by sitting at home.
With a machine Learning model using a Kaggle dataset to predict whether a person has a heart disease or not.
The Machine learning model was built using the scikit learn library available in python. The model was then stored using pickle. Data was collected from a user through a website whose backend was supported in flask.
- It was only my second time attempting to use flask, and I was overwhelmed with the number of tutorials available online and all of them saying difference syntaxes.
- I ran into few problems while coding but due to timezone differences with the mentors, they couldnt be cleared up and took longer than necessary to tackle
Incorporating particle.js in our website and deploying the machine learning model on the website using flask. Managing to find work arounds when there was technical failure.
So much. First of all I learnt to tackle college assignment deadline and hackathon deadline simultaenously. I learnt how to debug code. I understood how difficult it is to understand and make sense of someone elses code. Following tutorials in themselves is of no use, unless you understand the code and can make changes to suit your own needs.
- Two attempts were made to deploy the app on Google Cloud but they ended up being unsuccessful.
- Wanted to get my hands dirty with HTML and CSS in more detail but time was a constraint
- Improving the accuracy of the model by using deep learning models.
- Incorporating another data set which uses X-Ray scans to predict risk of heart diseases. More information could mean better accuracy of models.
- Working on the UI design of the website and making it more interactive and informative
- Deploying it on a website as well as making it available through a mobile app
- Deploying on GCP Tutorial followed: https://www.youtube.com/watch?v=RbejfDTHhhg
- Deploy Machine Learning model Flask: https://www.youtube.com/watch?v=i3RMlrx4ol4
- Underwater Scene Animation Effects using Particle.js: https://www.youtube.com/watch?v=LV8rYCmbmOo