Finding optimal learning rates with the Learning Rate Range Test

Learning Rates are important when configuring a neural network. But choosing one is not easy, as there is no single...

How to save and load a model with Keras?

So far, at MachineCurve, we have primarily focused on how to train models with Keras. This is nice, but a...

Convolutional neural networks are great tools for building image classifiers. They have been used thoroughly since the 2012 deep learning...

Sometimes, you don’t want the shape of your convolutional outputs to reduce in size. Other times, you wish to append...

Training Convolutional Neural Networks means that your network is composed of two separate parts most of the times. The last...

Sometimes, your data set may consist of e.g. images from which you only need to use a tiny bit in...

Creating ConvNets often goes hand in hand with pooling layers. More specifically, we often see additional layers like max pooling,...

There are three widely known regularization techniques for neural networks: L1 (or Lasso) regularization, L2 (or Ridge) regularization and Elastic...

The neurons of neural networks perform operations that are linear: they multiple an input vector with a weights vector and...

The inputs to individual layers in a neural network can be normalized to speed up training. This process, called Batch...