TrainDimensionalityReduction

Train a dimensionality reduction model

Description

Trainer for dimensionality reduction algorithms (autoencoders, PCA, SOM). All input samples are used to compute the model, like other machine learning models. The model can be used in the ImageDimensionalityReduction and VectorDimensionalityReduction applications.

Parameters

Input and output data

This group of parameters allows setting input and output data.

Input Vector Data -io.vd vectorfile Mandatory
Input geometries used for training (note: all geometries from the layer will be used)

Output model -io.out filename [dtype] Mandatory
Output file containing the estimated model (.txt format).

Input XML image statistics file -io.stats filename [dtype]
XML file containing mean and variance of each feature.

Field names to be used for training -feat string1 string2... Mandatory
List of field names in the input vector data used as features for training.

algorithm to use for the training -algorithm [som|autoencoder|pca] Default value: som
Choice of the dimensionality reduction algorithm to use for the training.

  • OTB SOM
    This group of parameters allows setting SOM parameters.
  • Shark Autoencoder
    This group of parameters allows setting Shark autoencoder parameters.
  • Shark PCA
    This group of parameters allows setting Shark PCA parameters.

OTB SOM options

Map size -algorithm.som.s string1 string2... Default value: 10 10
Sizes of the SOM map (one per dimension). For instance, [12;15] means a 2D map of size 12x15. Support2D to 5D maps.

Neighborhood sizes -algorithm.som.n string1 string2... Default value: 3 3
Sizes of the initial neighborhood in the SOM map (one per dimension). The number of sizes should be the same as the map sizes

NumberIteration -algorithm.som.ni int Default value: 5
Number of iterations for SOM learning

BetaInit -algorithm.som.bi float Default value: 1
Initial learning coefficient

BetaFinal -algorithm.som.bf float Default value: 0.1
Final learning coefficient

InitialValue -algorithm.som.iv float Default value: 10
Maximum initial neuron weight

Shark Autoencoder options

Maximum number of iterations during training -algorithm.autoencoder.nbiter int Default value: 100
The maximum number of iterations used during training.

Maximum number of iterations during training -algorithm.autoencoder.nbiterfinetuning int Default value: 0
The maximum number of iterations used during fine tuning of the whole network.

Epsilon -algorithm.autoencoder.epsilon float Default value: 0
Epsilon

Weight initialization factor -algorithm.autoencoder.initfactor float Default value: 1
Parameter that control the weight initialization of the autoencoder

Size -algorithm.autoencoder.nbneuron string1 string2... Mandatory
The number of neurons in each hidden layer.

Strength of the regularization -algorithm.autoencoder.regularization string1 string2... Mandatory
Strength of the L2 regularization used during training

Strength of the noise -algorithm.autoencoder.noise string1 string2... Mandatory
Strength of the noise

Sparsity parameter -algorithm.autoencoder.rho string1 string2... Mandatory
Sparsity parameter

Sparsity regularization strength -algorithm.autoencoder.beta string1 string2... Mandatory
Sparsity regularization strength

Learning curve -algorithm.autoencoder.learningcurve filename [dtype]
Learning error values

Shark PCA options

Dimension of the output of the pca transformation -algorithm.pca.dim int Default value: 10
Dimension of the output of the pca transformation.

Available RAM (MB) -ram int Default value: 256
Available memory for processing (in MB).

Examples

From the command-line:

otbcli_TrainDimensionalityReduction -io.vd cuprite_samples.sqlite -io.out model.som -algorithm som -algorithm.som.s 10 10 -algorithm.som.n 3 3 -algorithm.som.ni 5 -algorithm.som.bi 1 -algorithm.som.bf 0.1 -algorithm.som.iv 10 -feat value_0 value_1 value_2 value_3 value_4 value_5 value_6 value_7 value_8 value_9

From Python:

import otbApplication

app = otbApplication.Registry.CreateApplication("TrainDimensionalityReduction")

app.SetParameterString("io.vd", "cuprite_samples.sqlite")
app.SetParameterString("io.out", "model.som")
app.SetParameterString("algorithm","som")
app.SetParameterStringList("algorithm.som.s", ['10', '10'])
app.SetParameterStringList("algorithm.som.n", ['3', '3'])
app.SetParameterInt("algorithm.som.ni", 5)
app.SetParameterFloat("algorithm.som.bi", 1)
app.SetParameterFloat("algorithm.som.bf", 0.1)
app.SetParameterFloat("algorithm.som.iv", 10)
app.SetParameterStringList("feat", ['value_0', 'value_1', 'value_2', 'value_3', 'value_4', 'value_5', 'value_6', 'value_7', 'value_8', 'value_9'])

app.ExecuteAndWriteOutput()