Currently I am trying to work on an artificial neural network. I am also a very beginner in this field. Can any one help me giving any kind of helpful source, idea, or anything to apply this subject into the field of the strength of concrete beam?
Well, the first step for you is to look for previous research/papers on the application of ANN to predict the strength of concrete. Most importantly you should read and familiarize yourself with the concept of ANN and its applications and limitations. Also, if you do a search in the internet for "application of ANN in concrete strength" it will come up with useful results.
To start with, you can consult the following articles/references:
Dear Kumar: I have some experimental data relating a beam's depth, width, span, reinforcement area, concrete strength, weight etc. Now I am trying to utilize these data in the field of ANN to predict the strength a beam under a given circumstance. I want to know that can I use ANN on this regard? If so then what may be the easier or wiser way to approach. Also I am novice in ANN. Therefore I need help or advice, or suggestion regarding this.
Well, the first step for you is to look for previous research/papers on the application of ANN to predict the strength of concrete. Most importantly you should read and familiarize yourself with the concept of ANN and its applications and limitations. Also, if you do a search in the internet for "application of ANN in concrete strength" it will come up with useful results.
To start with, you can consult the following articles/references:
at the begining I sujest you to view web site: www.videolectures.net and search for information theory in events,1 event is by david mackay ,this event contain 16 lectures,3 of them is about ANN, as much as I know ANN is an algorithm for adjustment and prediction
You can use matlan software to develop a predictive model. You can use its ANN toolbox and also develop your own code. you should make a network using net command and then train your network using newff command. eventually you can simulate new data. It is so easy. you can type help newff to see matlab help. It is better that you randomized your data before dividing them into two different sets; training set and testing set. If you need more help I am ready to help you.