Researcher

Description

The Researcher class serves as the high-level API that integrates the core research modules through either inheritance or composition. This design provides flexibility and a structured workflow for users.

The Researcher class integrates the following submodules (as listed above):

Depending on the classification type, the implementation differs. The framework provides two specialized classes for binary and multi-class classification, namely BinaryResearcher and MultiClassResearcher. For simplicity, the name Researcher is used to refer to both classes in this documentation.

Main Class

class Researcher

A high-level API that integrates core research modules.

The Researcher class simplifies machine learning workflows by managing data handling, preprocessing, training, visualization, and experimentation in a unified interface.

__init__()

Initializes the Researcher class by integrating all research modules.

run_experiment(*args, **kwargs)

Conducts and manages machine learning experiments using the Experimenting module.

Parameters:

  • args (tuple) – Additional arguments for experiment configuration.

  • kwargs (dict) – Keyword arguments for defining experiment parameters.

Returns:

(Experiment) The experiment instance.

apply_preprocessing_pipeline(*args, **kwargs)

Applies a sequence of preprocessing steps using the Preprocessing module.

Parameters:

  • args (tuple) – Additional arguments for preprocessing configuration.

  • kwargs (dict) – Keyword arguments defining preprocessing parameters.

Returns:

(Any) The processed dataset after applying the preprocessing pipeline.

Class Methods

For recap, here are all the methods included in the Researcher class:

For binary classification, the following plotting methods are available:

For multi-class classification, the following plotting methods are available: