Deep Learning using Matlab

Deep Learning using Matlab - In this lesson, we will learn how to train a deep neural network using Matlab. It is divided into three sections -

1) Challenges of Deep Learning (continuation of previous lecture)

2) Designing the Architecture - https://youtu.be/HSDyiCuViWg?t=371

3) Writing the code in Matlab (Implementation) - https://youtu.be/HSDyiCuViWg?t=678

You can copy the code from here - https://www.nzfaruqui.com/deep-learning-using-matlab/

In the first section, the challenges of Deep Learning have been explained briefly. In the second section, the architecture of neural network has been explained. It also covered selection criteria of input nodes and output nodes, weight matrices and training data. In the third section, the implementation has been demonstrated.

In the previous lecture, we have learned details about Neural Network. That is why I didn't recap those things and directly started with the challenges of deep neural network.

Almost every new learners find it difficult to understand few things such as - how to select number of input nodes, number of output nodes, weight matrices, who data are passed into the network, how to calculate error and so on. I have explained these concept using animations and graphics.

Third section, the final section is the implementation part. Here I have implemented the concept I explained in the lecture. First I coded 'DeepLearning' function. The necessary functions - ReLU and Softmax function have been implemented as well. Then I have shown how to train a deep neural network by calling the 'DeepLearning' function. Finally I have tested the deep neural network. All of these has been shown in Matlab code.

I hope this lesson will help you to understand how to train a deep neural network in Matlab. And after completing this lecture, you will be able to train your own deep neural network to accomplish what ever you want.