CIFAR-10
Description
The CIFAR-10 dataset is a collection of images that are commonly used to train machine learning and computer vision algorithms. The CIFAR-10 dataset contains 60,000 32x32 color images in 10 different classes. There are 50,000 training images and 10,000 test images.
Neural network
This is the test with the neural network architecture used to obtain up to 36.80% accuracy on this dataset. To reach this accuracy you may need more attempts and more learning time.
TEST_F(Cifar10Test, trainNeuralNetwork)
{
StraightforwardNeuralNetwork neuralNetwork({
Input(3072),
FullyConnected(200),
FullyConnected(80),
FullyConnected(10)
});
neuralNetwork.train(*data, 1_ep || 300_s);
auto accuracy = neuralNetwork.getGlobalClusteringRate();
ASSERT_ACCURACY(accuracy, 0.24);
}