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 Supervised Learning Technique.
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 Spam Filtering with SVM (Example 3).
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 Statistical Model API.
This technique extends it with the following field:
-
SVC.
image
Image¶ Auto-generated Image if available
The implementation uses scikit-learn as Engine, there is no need of setting more than 1 Engine Meta Iterations
.
Model Parameters¶
-
SVC.
kernel
SVM Kernel¶ Kernel to be used in the SVM. If none is given, RBF will be used.
-
SVC.
penalty_parameter
Penalty parameter (C) of the error term.¶ Penalty parameter (C) of the error term.
-
SVC.
kernel_poly_degree
Polynomial Kernel degree¶ Degree of the Polynomial Kernel function. Ignored by all other kernels.
-
SVC.
kernel_coefficient
Kernel coefficient¶ Kernel coefficient for RBF, Polynomial and Sigmoid. Leave blank “for automatic” (1/n_features will be used)
-
SVC.
kernel_independent_term
Kernel Independent Term¶ Independent term in kernel function. It is only significant in Polynomial and Sigmoid kernels.
-
SVC.
class_weight
Class Weight¶ Set the parameter C of class i to class_weight[i]*C for SVC. If not given, all classes are supposed to have weight one. The “balanced” mode uses the values of y to automatically adjust weights inversely proportional to class frequencies in the input data as n_samples / (n_classes * np.bincount(y))
Implementation Parameters¶
-
SVC.
decision_function_shape
Decision Function Shape¶ Whether to return a one-vs-rest (‘ovr’) decision function of shape (n_samples, n_classes) as all other classifiers, or the original one-vs-one (‘ovo’) decision function of libsvm which has shape (n_samples, n_classes * (n_classes - 1) / 2).
-
SVC.
estimate_probability
Estimate Probability?¶ Whether to enable probability estimates. This will slow model fitting.
-
SVC.
use_shrinking
Use Shrinking Heuristic?¶ Whether to use the shrinking heuristic.
-
SVC.
tolerance
Tolerance¶ Tolerance for stopping criterion.
-
SVC.
cache_size
Kernel Cache Size (MB)¶ Specify the size of the kernel cache (in MB).
-
SVC.
random_seed
¶ The seed of the pseudo random number generator to use when shuffling the data. If int, random_state is the seed used by the random number generator; If RandomState instance, random_state is the random number generator; If None, the random number generator is the RandomState instance used by np.random.
-
SVC.
verbose
Be Verbose?¶ Enable verbose output. Note that this setting takes advantage of a per-process runtime setting in libsvm that, if enabled, may not work properly in a multithreaded context.