The network will have a configurable number of layers, activation functions on each layer as well as cost functions. Following posts will get a bit deeper into details such as choosing the activation functions, network architecture, weights initialization etc. The next post will explain and implement the backpropagation training algorithm in the same way. You will also need basic knowledge of linear algebra (matrix multiplication) and derivatives in order to understand the process in depth. It is designed for people who already have some coding experience as well as a basic understanding of what neural networks are and want to get a bit deeper into the way they work. This post will guide you through the process of building your own feed-forward multilayer neural network in Matlab in a (hopefully) simple and clean style.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |