Please follow me on Facebook: / nzamanfaruqui
Connect with me on LinkedIn: / nuruzzaman-faruqui
Subscribe to my new channel: / @nuruzzamanfaruqui
Contact me on Fiverr: www.fiverr.com/zaman_faruqui
In this lecture we will learn about single layer neural network. In order to learn deep learning, it is better to start from the beginning. And single layer neural network is the best starting point. This lecture starts with theoretical explanation covering only the essential elements of a neural network. This theoretical section is actually the foundation of leaning deep learning. That is why, it is very important to focus on this section and understand properly. After the theoretical section, a single layer neural network using SGD method has been trained in Matlab. In a nutshell, this lecture covers -
1) Concept of Nodes
2) Concept of Layers
3) Supervised Learning
4) Delta Rule and Generalize Delta Rule
5) SGD, Batch and Mini Batch Method
6) Practical Implementation (Training a Single Layer Neural Network in Matlab)
Most of the time in university class environment students learn the theoretical aspects only. To be honest, it is difficult to cover both theorical and practical aspects in the class. However, this is the best approach of learning anything. And of course, the best of anything comes with a price. I’ve noticed that very few university instructors of my country have the mindset of teaching in proper way - which is the combination of theory and implementation of the theory. Learning without gaining the ability to apply is useless. If you are a victim of these types of poor education system, this lecture will help you to learn neural network properly.
After following this lecture properly, a student will be able to implement single layer neural network in Matlab. The example shown here is done in Matlab. It does not mean that you are bound to use Matlab. Using the concept explained and procedure shown here, you can train single layer neural network using the programming language and IDE you are comfortable with.
22 фев 2019
