Deep Learning with Microsoft Cognitive Toolkit Quick Start Guide

Checking the performance of the neural network

Every time we pass data through the trainer to optimize our model, it measures the performance of the model through the metric that we configured for the trainer. The model performance measured during training is on the training set. It is useful to measure the accuracy on the training set, because it will tell you whether the model is actually learning anything from the data.

For a full analysis of the model performance, you need to measure the performance of the model using the test set. This can be done by invoking the test_minibatch method on the trainer as follows:

trainer.test_minibatch( {features: X_test, label: y_test })

This method accepts a dictionary with a mapping between the input ...