sits (version 0.13.0)

sits_TempCNN: Train a model using the Temporal Convolutional Neural Network

Description

Use a TempCNN algorithm to classify data, which has two stages: a 1D CNN and a multi-layer perceptron. Users can define the depth of the 1D network, as well as the number of perceptron layers.

This function is based on the paper by Charlotte Pelletier referenced below and code available on github (https://github.com/charlotte-pel/temporalCNN). If you use this method, please cite the original tempCNN paper.

Usage

sits_TempCNN(
  samples = NULL,
  cnn_layers = c(64, 64, 64),
  cnn_kernels = c(5, 5, 5),
  cnn_activation = "relu",
  cnn_L2_rate = 1e-06,
  cnn_dropout_rates = c(0.5, 0.5, 0.5),
  mlp_layers = c(256),
  mlp_activation = "relu",
  mlp_dropout_rates = c(0.5),
  optimizer = keras::optimizer_adam(lr = 0.001),
  epochs = 150,
  batch_size = 128,
  validation_split = 0.2,
  verbose = 0
)

Arguments

samples

Time series with the training samples.

cnn_layers

Number of 1D convolutional filters per layer

cnn_kernels

Size of the 1D convolutional kernels.

cnn_activation

Activation function for 1D convolution. Valid values: 'relu', 'elu', 'selu', 'sigmoid'.

cnn_L2_rate

Regularization rate for 1D convolution.

cnn_dropout_rates

Dropout rates for 1D convolutional filters.

mlp_layers

Number of nodes in the multi-layer-perceptron.

mlp_activation

Names of 2D activation functions for the MLP. Valid values: 'relu', 'elu', 'selu', 'sigmoid'.

mlp_dropout_rates

Dropout rates (0,1) for each layer in the MLP.

optimizer

Function with a pointer to the optimizer function (default is optimization_adam()). Options: optimizer_adadelta(), optimizer_adagrad(), optimizer_adam(), optimizer_adamax(), optimizer_nadam(), optimizer_rmsprop(), optimizer_sgd().

epochs

Number of iterations to train the model.

batch_size

Number of samples per gradient update.

validation_split

Number between 0 and 1. Fraction of training data to be used as validation data. The model will set apart this fraction of the training data, will not train on it, and will evaluate the loss and any model metrics on this data at the end of each epoch. The validation data is selected from the last samples in the x and y data provided, before shuffling.

verbose

Verbosity mode (0 = silent, 1 = progress bar, 2 = one line per epoch).

Value

A fitted model to be passed to sits_classify

References

Charlotte Pelletier, Geoffrey Webb and Fran<U+00E7>ois Petitjean, "Temporal Convolutional Neural Network for the Classification of Satellite Image Time Series", Remote Sensing, 11,523, 2019. DOI: 10.3390/rs11050523.

Examples

# NOT RUN {
# Retrieve the set of samples for the Mato Grosso (provided by EMBRAPA)

# Build a machine learning model based on deep learning
tc_model <- sits_train(samples_modis_4bands, sits_TempCNN(epochs = 75))
# Plot the model
plot(tc_model)

# get a point and classify the point with the ml_model
point <- sits_select(point_mt_6bands, bands = c("NDVI", "EVI", "NIR", "MIR"))
class <- sits_classify(point, tc_model)

plot(class, bands = c("NDVI", "EVI"))
# }