Skip to content

Latest commit

 

History

History
71 lines (53 loc) · 2.95 KB

supervised_learning.rst

File metadata and controls

71 lines (53 loc) · 2.95 KB
Raw

Supervised Learning

This app provides Supervised Learning techniques for integrating them into systems or directly to your code.

From an API point of view, each technique is a particular implementation of :ref:`api_supervised_learning`.

Support Vector Machines (SVM)

Support Vector Machines are provided by integrating the scikit-learn framework: http://scikit-learn.org.

If you are not familiar with the framework, it is better at least take a glance on its excellent documentation for the technique for a better understanding on how the modelling is done.

An example of integrating SVM into a system can be found in :ref:`example_spam_filtering`.

SVM for Classification

All the configuration can be done through the admin of Support Vector Machines for Classification - or more specifically, through the change form.

The following fields are available for configuration:

General

General fields (like Name) and Miscellanous are documented in the :ref:`api_statistical_model`.

This technique extends it with the following field:

.. autoattribute:: supervised_learning.models.svm.SVC.image
    :annotation: Image

The implementation uses scikit-learn as Engine, there is no need of setting more than 1 Engine Meta Iterations.

Model Parameters

.. autoattribute:: supervised_learning.models.svm.SVC.kernel
    :annotation: SVM Kernel

.. autoattribute:: supervised_learning.models.svm.SVC.penalty_parameter
    :annotation: Penalty parameter (C) of the error term.

.. autoattribute:: supervised_learning.models.svm.SVC.kernel_poly_degree
    :annotation: Polynomial Kernel degree

.. autoattribute:: supervised_learning.models.svm.SVC.kernel_coefficient
    :annotation: Kernel coefficient

.. autoattribute:: supervised_learning.models.svm.SVC.kernel_independent_term
    :annotation: Kernel Independent Term

.. autoattribute:: supervised_learning.models.svm.SVC.class_weight
    :annotation: Class Weight

Implementation Parameters

.. autoattribute:: supervised_learning.models.svm.SVC.decision_function_shape
    :annotation: Decision Function Shape

.. autoattribute:: supervised_learning.models.svm.SVC.estimate_probability
    :annotation: Estimate Probability?

.. autoattribute:: supervised_learning.models.svm.SVC.use_shrinking
    :annotation: Use Shrinking Heuristic?

.. autoattribute:: supervised_learning.models.svm.SVC.tolerance
    :annotation: Tolerance

.. autoattribute:: supervised_learning.models.svm.SVC.cache_size
    :annotation: Kernel Cache Size (MB)

.. autoattribute:: supervised_learning.models.svm.SVC.random_seed
    :annotation:

.. autoattribute:: supervised_learning.models.svm.SVC.verbose
    :annotation: Be Verbose?