Learning algorithm for multi-layered networks, Computer Engineering

Learning algorithm for multi-layered networks:

Furthermore details we see that  if S is too high, the contribution from wi * xi is reduced. It means that t(E) - o(E) is multiplied by xi after then if xi is a big value as positive or negative so the change to the weight will be greater. Here to get a better feel for why this direction  correction works so it's a good idea to do some simple calculations by hand. 

Here η simply controls how far the correction should go at one time that is usually set to be a fairly low value, e.g., 0.1. However the weight learning problem can be seen as finding the global minimum error which calculated as the proportion of mis-categorised training examples or over a space when all the input values can vary. Means it is possible to move too far in a direction and improve one particular weight to the detriment of the overall sum: whereas the sum may work for the training example being looked at and it may no longer be a good value for categorising all the examples correctly. Conversely for this reason here η restricts the amount of movement possible. Whether large movement is in reality required for a weight then this will happen over a series of iterations by the example set. But there sometimes η is set to decay as the number of that iterations through the entire set of training examples increases it means, can move more slowly towards the global minimum in order not to overshoot in one direction.

However this kind of gradient descent is at the heart of the learning algorithm for multi-layered networks that are discussed in the next lecture. 

Further Perceptrons with step functions have limited abilities where it comes to the range of concepts that can be learned and as discussed in a later section. The other one way to improve matters is to replace the threshold function into a linear unit through which the network outputs a real value, before than a 1 or -1. Conversely this enables us to use another rule that called the delta rule where it is also based on gradient descent.

Posted Date: 1/11/2013 7:02:52 AM | Location : United States







Related Discussions:- Learning algorithm for multi-layered networks, Assignment Help, Ask Question on Learning algorithm for multi-layered networks, Get Answer, Expert's Help, Learning algorithm for multi-layered networks Discussions

Write discussion on Learning algorithm for multi-layered networks
Your posts are moderated
Related Questions
Q. Data Alignment in Arrays? Arrays are aligned to templates by ALIGN directive. ALIGN directive is used to align elements of various arrays with one another denoting that they

Interconnection Network Introduction This unit discusses the types and properties of interconnection networks.  In multiprocessor systems, there are multiple I/O modules

Utilization Summary The Utilization Summary shows the status of each processor i.e. how much time (in the form of percentage) have been spent by every processor in busy mode, o

Video Conferencing Video conferencing continues to grow in popularity. Why is this? Some reasons are listed below: - Communication links are now much faster thus sound quali

Q. Illustrate Character Coded Data? The input output happens in the form of ASCII data. These ASCII characters are entered as a string of data. For illustration to get two numb

Create a new project, LastNameFirst 05, for this assignment. You will need to define the classes depicted in the class diagram below. All of the classes should contain complete ja

Spirit Duplicating Equipment Required Spirit Duplicator (also known as hectograph) Thermal copier (optional) Materials Masters Hectographic carbon COPY pa

What is Directory?   The device directory or simply called as directory records information-such as name, location, size, and type for all files on that certain partition. The

What courses draught in a simple pendulum experiment

how to become an ict enginer