An android app with a python backend to do diabetes prediction
The android app uses a python backend to do the prediction based on an aggregation of models. The results and data are written to a database. The app connects to a server through a URL. A base request handler handles the request here, through a socket. Then, it calls a Classification function. The results of the classification function are sent to the android app via callback function. Then, the results are written to the database with a user identifier.
The function returns the aggregate result of several models, whose results are summarized as text and returned.
The callback function returns to the app first a response of 200. If there is no result then a 400 is returned. This is followed by a string containing the results.
The app displays an error and handles appropriately.
The app receives the result and displays it.
There is a splash screen at first. Then the app has a login/registration id page, along with server address. This is used to login. The app then stores the data from the user. It sends the data to the server. It displays a result.
use flask instead of the base request handler: Search for 'simple flask app' in google.
Attach a weight tracker to the app.
Display list of improvements in health the user can make.
-
- Train the models and save them as files.
- Load and test them for predictions.
- Create a class that classifies input data.
- Summarize the aggregate result of several models as text in the class and analyse accuracy.
- Return the accuracy and the predictions.
- Test it with stubs.
-
- Create a nav drawer page as the main page.
- Create the splash screen and test it with stubs.
- Create a login and registration page and test it with stubs.
- Create a form page for user data input
- Create a result page and test it with stubs.
-
- Create a flask app. If that doesn't work, create a simple HTTP base handler server.
- Set it to handle some requests.
- Test the requests with stubs.
- Set it to return responses.
- Test the responses with stubs.
- Create the Classification function.
- Attach it to the Classification function.
- Test requests and responses with stubs.
-
- Attach the server to the app.
- Test for sample input data.
- Remove bugs.
- Publish as Alpha version.
-
- Test it for a wide range of input data.
- Find its limitations.
- Document it's expected input and output.
-
- Create the project report.
- Use README.md as a reference to create the report introduction and methodology.
- Display statistics and figures about the data.
- Show mockup figures.
- Display expected results for sample inputs.
- Talk about future imporvements.
- Conclusion
- Create the project PowerPoint.
- Turn the report into a PPT.
- Divide roles of speaking.
- Create a paper on the software.
- Create the project report.