Autokeras: GeneralRegressor

Created on 22 Aug 2018  路  6Comments  路  Source: keras-team/autokeras

Suggested Name

Regression

Task Description

A general regression problem with both numeric and categorical features. The input is tabular data.
The regression target is a single value for each input data which is a row in the table.

Evaluation Metrics

Mean Square Error

Benchmark Datasets

NA

Reason

There are a lot of regression problems in machine learning

Solution

NA

Additional Context

NA

Most helpful comment

I can not find the GeneralRegressor in the documentation, or in the code examples

All 6 comments

130 Thanks for the request. I'm currently working on these two issues.

This issue is solved. Please see the code in the examples directory.

Does this rely on lightgbm only or cover DL architectures as well?

I can not find the GeneralRegressor in the documentation, or in the code examples

can't find this either, seems to be issues with GeneralRegressor, TabularClassifier etc. but nothing in the codebase or docs.

can't find this either, seems to be issues with GeneralRegressor, TabularClassifier etc. but nothing in the codebase or docs.

I can not find the GeneralRegressor in the documentation, or in the code examples

AutoKeras Tabular was removed, reason for this is explained in #562

Thankfully @bolkedebruin created a package (https://pypi.org/project/autokeras-tabular/) containing the missing functionality.

Was this page helpful?
0 / 5 - 0 ratings