datamllab/autokaggle

Automated Machine Learning (AutoML) for Kaggle Competition

autokaggle's Issues

Change the example from autokeras to autokaggle

Enable the coverall and codacy

Benchmarks

Run our code on the kaggle datasets.
Summarize the results in a table.
The human solution performances.

Is the project ready to be tested?

I am checking the readme and the code. and it looks like this project is not prepared for testing would that be correct? I say so because the examples show

from autokaggle import TabularClassifier

Which fails to import (from autokaggle.tabular_supervised import TabularClassifier, would work), I assume the readme was written before the api is designed.It looks like an awesome project and want to make sure its ready :)

Enable Pypi deployment

Automatic GitHub pages

Auto push to gh-pages branch after build in CI.

Design the API

Should be written in GitHub gist.
It should be a code example of user calling the APIs.
The code should contain comments.

The new API reserves the current functionality,
but the automated feature engineering can be called independently.