honcaml 0.1.0

Holistic and No Code Auto Machine Learning

HoNCAML

HoNCAML (Holistic No Code Automated Machine Learning) is a tool aimed to run automated machine learning pipelines for problems of different nature; main types of pipeline would be:

1. Training the best possible model for the problem at hand
2. Use this model to predict other instances

Why HoNCAML

Focus

HoNCAML has been designed having the following aspects in mind:

• Ease of use
• Modularity
• Extensibility
• Simpler is better

Users

There are (at least) two main types of users who could benefit from this tool:

1. Regular users: In terms of programming experience and/or machine learning knowledge. It would be possible for them to get results in an easy way.
2. Advanced users: It is possible to customise experiments in order to adapt to a specific use case that a user with previous knowledge would like.

Pipelines

This library assumes data has tabular format, and is clean enough to be used to train models.

At this moment, the following types of problems are supported:

• Regression
• Classification

Regarding available models, the following are supported:

• Sklearn models
• Pytorch (neural net) models

However, due to its nature, extend the library to include other type of problems and models should be not only feasible, but intuitive.

Installation

To set up and install HoNCAML, just run the following within a virtual environment:

make install

Virtual environment directory is located in ./venv by default, but it can be changed by changing the variable ENV_PATH located in Makefile.

Quick usage

For a quick train execution, given that a dataset is available with the target value informed, it is necessary to first create a basic configuration file:

honcaml -b {config_file} -t {pipeline_type}

Being {config_file} the path to the file containing the configuration in yaml extension, and being {pipeline_type} one of the supported: train, predict or benchmark.

The specified keys of the file should be filled in, and afterwards it is possible to run the intended pipeline with the following command:

honcaml -c {config_file}

This will run the pipeline and export the trained model.

Detailed configuration

In the case of advanced configuration, there is the option of generating a more complete one, instead of the basic mentioned above:

   honcaml -a {config_file} -t {pipeline_type}

Advanced configuration files contain comments with required information to fill in the blanks. All the details of the configuration file are explained in the documentation. Moreover, many examples can be found at examples.

Executing from the GUI

To run the HoNCAML GUI locally in a web browser tab, run the following command:

honcaml -g

It allows to execute HoNCAML providing a datafile and a configuration file, or to manually select the configuration options instead of providing the file.

When using the manual configuration, it allows both levels of configuration: Basic, for a faster execution, and Advanced, allows users to configure the model hyperparameters; and three functionalities: Benchmark, Train and Predict.

Contribute

All contributions are more than welcome! For further information, please refer to the contribution documentation.

Bugs

If you find any bug, please report it as an issue.

Contact

Should you have any inquiry regarding the library or its development, please contact the Applied Machine Learning team.