The fashion and this data set was created by [inaudible] and [inaudible]. I think it's really cool that you're already able to implement a neural network to do this fashion classification task. It's just amazing that large data sets like this are readily available to students so that they can learn and it make it really easy to learn. And in this case we saw with just a few lines of code, we were able to build a DNN that allowed you to do this classification of clothing and we got reasonable accuracy with it but it was a little bit of a naive algorithm that we used, right? We're looking at every pixel in every image, but maybe there's ways that we can make it better but maybe looking at features of what makes a shoe a shoe and what makes a handbag a handbag. What do you think? Yeah. So one of the ideas that make these neural networks work much better is to use convolutional neural networks, where instead of looking at every single pixel and say, "Oh, that pixel has value 87, that has value 127." So is this a shoe or is this a hand bag? I don't know. But instead you can look at a picture and say, "Oh, I see shoelaces and a sole." Then, it's probably shoe or say, "I see a handle and rectangular bag beneath that." Probably a handbag. So confidence hopefully, we'll let the students do this. Sure, what's really interesting about convolutions is they sound complicated but they're actually quite straightforward, right? It's a filter that you pass over an image in the same way as if you're doing sharpening, if you've ever done image processing. It can spot features within the image as you've mentioned. With the same paradigm of just data labels, we can let a neural network figure out for itself that it should look for shoe laces and soles or handles in bags and just learn how to detect these things by itself. So shall we see what impact that would have on Fashion MNIST? So in the next video, you'll learn about convolutional neural networks and get to use it to build a much better fashion classifier.