gssl.classifiers.nn package

Submodules

gssl.classifiers.nn.NN module

A Neural Network approach I was previously using for some experiments. Ignore this for now.

class gssl.classifiers.nn.NN.Accumulator(tensor, name, init_val=0.0)

Bases: object

__init__(tensor, name, init_val=0.0)

Initialize self. See help(type(self)) for accurate signature.

class gssl.classifiers.nn.NN.NNClassifier(image_shape=None, OUT_SIZE=10, NUM_EPOCHS=2000, BATCH_SIZE=50, AUGMENT=False, model_choice='simple', LEARNING_RATE=0.0001)

Bases: gssl.classifiers.classifier.GSSLClassifier

A NN classifier that was intended to optimize a labeled and unlabeled objective at the same time. Please ignore this for the time being

ALPHA = 0.1

LAMBDA = 0.5

RECALC_W = False

SIGMA = <tf.Tensor: id=0, shape=(), dtype=float32, numpy=0.04>

USE_UNLABELED = True

__init__(image_shape=None, OUT_SIZE=10, NUM_EPOCHS=2000, BATCH_SIZE=50, AUGMENT=False, model_choice='simple', LEARNING_RATE=0.0001)

Constructor for NN classifier.

Args:

build_graph(X, k)

eval_get_data = None

evaluate_simfunc(W_sparse_vals)

fit(X, W, Y, labeledIndexes, hook=None, Y_true=None)

Classifies the input data.

Parameters

• X (NDArray[float].shape[N,D]) – Input matrix of N instances of dimension D.

• W (NDArray[float].shape[N,N]) – The affinity matrix encoding the weighted edges.

• Y (NDArray[float].shape[N,C]) – The initial belief matrix

• hook (GSSLHook) – Optional. A hook to execute extra operations (e.g. plots) during the algorithm

Returns

An updated belief matrix.

Return type

NDArray[float].shape[N,C]

labeled_gen()

pred_gen()

random_gen()

unlabeled_gen()

unlabeled_pairs_gen()

gssl.classifiers.nn.NN.convert_sparse_matrix_to_sparse_tensor(X, var_values=False)

gssl.classifiers.nn.NN.cos_decay(init_val, EPOCH_VAR, rampdown_length)

gssl.classifiers.nn.NN.debug(msg)

gssl.classifiers.nn.NN.ent(Y)

gssl.classifiers.nn.NN.gather(x, F)

gssl.classifiers.nn.NN.get_S(W)

gssl.classifiers.nn.NN.get_S_fromtensor(W)

gssl.classifiers.nn.NN.kl_divergence(self, p, q)

gssl.classifiers.nn.NN.repeat(x, n)

gssl.classifiers.nn.NN.row_normalize(x)

gssl.classifiers.nn.NN.xent(distr_1, distr_2)

gssl.classifiers.nn.models module

Created on 4 de out de 2019

@author: klaus

gssl.classifiers.nn.models.conv_large(input_shape, output_shape)

gssl.classifiers.nn.models.conv_small(input_shape, output_shape)

gssl.classifiers.nn.models.linear(input_shape, output_shape)

gssl.classifiers.nn.models.simple(input_shape, output_shape)

Module contents