MDS (DistMatrix)

This node maps data of a high dimensional space onto a lower (usually 2 or 3) dimensional space. Therefore the Sammons mapping is applied, which iteratively decreases the difference of the distances of high and low dimensional data. Each original data point is represented by a data point of a lower dimension. The Sammons mapping tries to keep the distance information of the high dimensional data by adjusting the low dimensional data points in a certain way. Each low dimensional data point is moved around a bit towards or back from the other points according to its high dimensional distances. This procedure is repeated a specified number of epochs or iterations respectively. The distances in the original high dimensional space have to be provided by a column containing distance vectors (see distance matrix calculate node). According to these distances the data is mapped onto the lower dimensional space. In the low dimensional space the Euclidean distance is used to arrange the points.

Options

Number of rows to use
Specifies the number of rows to apply the MDS on.
Output dimension
Specifies the dimension of the mapped output data.
Epochs
Specifies the number of epochs to train.
Learn rate
Specifies the learning rate to use. The learning rate is decreased automatically over the trained epochs.
Random seed
Specifies the random seed to use, which allows to reproduce a mapping even if the initialization is done randomly.
Distance matrix column
The column containing the distance vectors of the original high dimensional data to map.

Input Ports

Icon
Data table containing the data to map.

Output Ports

Icon
The input data and the mapped data.

Views

This node has no views

Workflows

Links

Developers

You want to see the source code for this node? Click the following button and we’ll use our super-powers to find it for you.