Image Classification Tutorial-3 (Fashion MNIST datasets using convolutional neural network)
Hello Friends, Welcome to the Automation Anomaly!!
In today's exciting tutorial we are going to train Convolutional Neural Network to classify MNIST Fashion datasets. This video is revision of our previous tutorial but we will learn the simple fine tuning method in this video.
Let me give you quick snapshot of today's tutorial so, In this video we are going to learn the architecture and brief concept of Convolutional Neural Networks. In this tutorial we will see the simple fine tuning techniques to get best accuracy. Basically, we will change the learning rate and we will change number of filters and also increase some layers to increase the accuracy.
Once you have understand all the basic concepts we are ready to train our Convolutional Neural Network to recognize the MNIST Fashion datasets. After training CNN we will check the training and validation accuracy and loss by plotting the data. Also, we have to check that how CNN will perform on the unseen test datasets.
So this tutorial is complete package for those who want to learn deep learning and neural network.
In my next tutorial we will make our first image classifier to identify Dog an cat. So from now onward we will take raw datasets from the internet and we will do image preprocessing. So in my next tutorial we will test on images as well and also we will see the very important concept of "Confusion Matrix" and "Classification Report" to check the model accuracy.
Thank you so much for stopping by at Automation Anomaly!
Have a fantastic day ahead.
Stay Safe Stay Healthy.