iota2.learning.pytorch.torch_nn_bank

Module containing all natively usable pytorch neural networks in iota2

Functions

abstractmethod(funcobj)	A decorator indicating abstract methods.
reduce(function, sequence[, initial])	Apply a function of two arguments cumulatively to the items of a sequence, from left to right, so as to reduce the sequence to a single value.
verify_import(module_path)	Check if a module can be imported

Classes

ABC()	Helper class that provides a standard way to create an ABC using inheritance.
ABCMeta(name, bases, namespace, **kwargs)	Metaclass for defining Abstract Base Classes (ABCs).
ANN(nb_features, nb_class[, layer, ...])	Same as ANN but reshape it's inputs
Iota2NeuralNetwork(nb_features, nb_class[, ...])	every neural networks provided by iota2 must inherit from this one to ensure some specific signatures.
Iota2NeuralNetworkFactory()	factory to a Iota2NeuralNetwork object
LTAE(*args, **kwargs)
LTAEClassifier(nb_features, nb_class[, ...])	SITS classifier using a transformer encoder
MLPClassifier(nb_features, nb_class[, ...])	Simple Multi Layer Perceptron for Satellite Image Time Series classification
MaskedAttention(nb_features, nb_bands)	Attention layer using validity masks as context vector
NeuralNetworkMeta(name, bases, namespace, ...)	metaclass which check every abstract methods definition
SelfAttention(input_size, output_size)	Self attention layer using dot product and addition
SimpleSelfAttentionClassifier(nb_features, ...)	SITS classifier using validity masks and self attention.