A demo for support vector regression

Demo displaying the performance of SVR in Shogun.
You can draw your own points or have some generated for you from the "Toy Data" panel.
You can also play around with the various parameters to
see how they effect the outcome.

A demo for ridge regression

Demo displaying the performance of various regression models in Shogun : Linear Ridge regression, Least Squares regression, Kernel ridge regression.
You can draw your own points or have some generated for you from the "Toy Data" panel where you can also load the Boston housing dataset.
You can also play around with the various parameters to
see how they effect the outcome.

A demo for gaussian process regression

Demonstration of performing regression with Gaussian Processes in Shogun.
A detailed how-to can be found in the GP notebook .

You can enter your own data points by left clicking on the canvas below or have some generated for you from the "Toy Data" panel on the right.

Use the learn parameters dropdown to automagically learn the best hyper-parameters of the GP using maximum likelihood (ML2).

You can set inducing points for FITCInferenceMethod by right clicking on the canvas.

A demo for binary classification

Demonstration of a binary classification task with Shogun, using the
CLibSVM
class.
You can enter instances of the red and blue classes by left and right-clicking on the canvas below or you can have some points generated for you from the 'Toy Data' panel on the right.You can also experiment with the various parameters on the right to see how they affect the outcome.

A demo for binary perceptron

Demonstration of binary classification with Shogun, using the
CPerceptron
class which will provide us with a linear classifier.
You can enter instances of the red and blue classes by left and
right-clicking on the canvas below.
You can also experiment with the various parameters on the right to see how they affect the outcome.

A demo for multiclass classification

Demonstration of multiclass classification with Shogun, using the
GMNPSVM
class.
You can select a class from the appropriate area below the canvas and then draw points for that class by clicking on the canvas
or have some data generated for you from the 'Toy Data' panel on the right. You can also experiment with the various parameters on the right to see how they affect the outcome.

A demo for gaussian process classification

Demonstration of binary classification with Shogun, using the
GaussianProcessBinaryClassification
class.
You can enter instances of the red and blue classes by left and right-clicking on the canvas below or you can have some points generated for you from the 'Toy Data' panel on the right.You can also experiment with the various parameters on the right to see how they affect the outcome.
To get best parameters using model selection try the learn parameters dropdown.

A demo for kernel matrix visualization

Click on the canvas below to enter some points and experiment with
the various arguments to see how the kernel matrix is affected each time.

You can also have some data points generated for you from the "Toy Data" panel on the right

A demo for recognizing hand-written digits.

This demo application uses a previously trained
CGMNPSVM
svm in combination with the
Gaussian Kernel
to recognize hand-written digits.
To test it, draw a digit (0..9) in the area below and press recognize!

A demo for language detection

This application demo uses a previously trained
MulticlassLibLinear
svm, in conjuction with the HashedDocDotFeatures, to predict the language of documents.

It works for 5 languages: English, Greek, German, Italian and Spanish.

A demo for clustering using kmeans

Clustering demonstration using the CKMeans class of Shogun.
More information on the k-means clustering algorithm can be found here.

You can enter data points by clicking on the canvas below or you can have some generated for you from the "Toy Data" panel on the right.

You can also experiment with the arguments to see how they affect the outcome.