Neural Networks and TensorFlow - Deep Learning Series [Part 22]

It's been a while (about two weeks) since I posted my last video in the series of deep learning with TensorFlow.

We've gone quite deep into the series, and it's gonna get even longer; can't wait to see it have maybe 50 lessons or even more.

In this particular lesson we're going to create our model of neural network (a convolutional neural network, which is well suited for image data and image processing). We're going to train this CNN on the CIFAR10 dataset.

The architecture of the net is quite common to the usual CNN: it is composed of the inputs (of course), convolutional layers, pooling layers, and fully connected layers.

Once this computation graph is constructed, we're going to create a TensorFlow session, initialize all variables, and then execute the graph so that we can train our network. But more about that later; here we're only constructing the graph. See the video for the full breakdown of the lesson.

To stay in touch with me, follow @cristi

Cristi Vlad Self-Experimenter and Author