Product Requirements - SD-Group-11/ml-frontend GitHub Wiki

The following is a list of initial high-level requirements as specified by the client for the end product:

• An application or web based interface that allows users to register and login to their accounts.
• There should be tutorials and help buttons that explain to the users how to operate the software.
• The ability for the user to train simple machine learning models such as linear regression, naive bayes and support vector machines.
• Users should be able to upload or read files that they would like to learn/fit the machine learning models to.
• Each user should only be able to view their own uploaded files, later functionality could possibly be added in order to allow users to make their data files public so that others can download it from a dataset section of the software.
• Tutorials should be available for each model to guide users in uploading the correct file format in order for the software to correctly read the uploaded file.
• Users should have the ability to tune their own hyper parameters for each of the simple machine learning models.
• There should be graphical plotting functionality within the software so that users can visualize how their models fit the data and present different outputs. This may allow us to later plot many types of graphs later like comparison and convergence plots.
• Users should be able to save those plots (in jpg, pdf, etc.) to their local drives.
• The software should allow users to save the trained models and also allow them to upload the save points so that they can use previously trained models and perform tests on them without retraining again.