MXNET: Weight attenuation-gluon implementation

Building a data set

# -*- coding: utf-8 -*-from mxnet import initfrom mxnet import ndarray as ndfrom mxnet.gluon import loss as glossimport gbn_train = 20n_test = 100num_inputs = 200true_w = nd.ones((num_inputs, 1)) * 0.01true_b = 0.05features = nd.random.normal(shape=(n_train+n_test, num_inputs))labels = nd.dot(features, true_w) + true_blabels += nd.random.normal(scale=0.01, shape=labels.shape)train_features, test_features = features[:n_train, :], features[n_train:, :]train_labels, test_labels = labels[:n_train], labels[n_train:]

Data iterators

from mxnet import autogradfrom mxnet.gluon import data as gdatabatch_size = 1num_epochs = 10learning_rate = 0.003train_iter = gdata.DataLoader(gdata.ArrayDataset(    train_features, train_labels), batch_size, shuffle=True)loss = gloss.L2Loss()

Train and Show results

Gb.semilogy function: loss to draw training and test data

From mxnet import gluonfrom mxnet.gluon import nndef fit_and_plot (Weight_decay): NET = nn. Sequential () Net.add (NN. Dense (1)) Net.initialize (init.    Normal (sigma=1) # L2 norm regularization of weight parameters, that is, weight attenuation. Trainer_w = gluon. Trainer (Net.collect_params ('. *weight '), ' sgd ', {' learning_rate ': learning_rate, ' WD ': Weight_decay}) # do not deviate parameters    Regularization of the L2 norm. Trainer_b = gluon.    Trainer (Net.collect_params ('. *bias '), ' sgd ', {' learning_rate ': learning_rate}) Train_ls = [] Test_ls = [] For _ in range (Num_epochs): For X, y in Train_iter:with Autograd.record (): L = loss (net            (X), y) L.backward () # Call the step function on the two Trainer instances, respectively.                             Trainer_w.step (batch_size) trainer_b.step (batch_size) train_ls.append (loss (NET (train_features),                            train_labels). mean (). Asscalar ()) Test_ls.append (loss (NET (test_features), test_labels). mean (). Asscalar ()) gb.semilogy (rangE (1, Num_epochs + 1), Train_ls, ' epochs ', ' loss ', range (1, Num_epochs + 1), Test_ls, [' Train ', ' Test ']) Return ' w[:10]: ', Net[0].weight.data () [:,: ten], ' B: ', Net[0].bias.data () print Fit_and_plot (5)

• Using Gluon's WD Hyper-parameters, you can use weight attenuation to cope with overfitting problems.
• We can define multiple Trainer instances to use different iterative methods for different model parameters.

MXNET: Weight attenuation-gluon implementation