https://github.com/bolkedebruin/autokeras_tabular
Autokeras Tabular extension
https://github.com/bolkedebruin/autokeras_tabular
Last synced: about 2 months ago
JSON representation
Autokeras Tabular extension
- Host: GitHub
- URL: https://github.com/bolkedebruin/autokeras_tabular
- Owner: bolkedebruin
- License: other
- Created: 2019-04-16T09:48:43.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2019-07-15T03:04:15.000Z (over 6 years ago)
- Last Synced: 2025-04-10T19:47:24.675Z (6 months ago)
- Language: Python
- Size: 17.6 KB
- Stars: 16
- Watchers: 4
- Forks: 3
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE.txt
Awesome Lists containing this project
README
# Autokeras Tabular
Tabular support was removed by the Autokeras authors without a replacement. This reinstates the module. It should
work exactly the same as before except you will need to import from autokeras_tabular.
### Automated tabular classifier tutorial. [[source]](https://github.com/bolkedebruin/autokeras_tabular/tree/master/examples/task_modules/tabular)
Class `TabularClassifier` and `TabularRegressor` are designed for automated generate best performance shallow/deep architecture
for a given tabular dataset. (Currently, theis module only supports lightgbm classifier and regressor.)
```python
clf = TabularClassifier(verbose=True)
clf.fit(x_train, y_train, time_limit=12 * 60 * 60, data_info=datainfo)
```
* x_train: string format text data
* y_train: int format text label
* data_info: a numpy.array describing the feature types (time, numerical or categorical) of each column in x_train.
**Notes:** Preprocessing of the tabular data:
* Class `[TabularPreprocessor]` involves several automated feature preprocessing and engineering operation for tabular data .
*The input data should be in numpy array format for the class `TabularClassifier` and `TabularRegressor` .