Guided Independent Study (GIS) Week (20/4/2009-24/4/2009)


During the Guided Independent Study (GIS) week you will have the opportunity to develop your skills of extending your knowledge by self study and study/search of appropriate sources, while practicing and experimenting with the new material you learn about.

During this week there will be no lectures. You are expected to study in an independent manner practice your implementation skills and learn more about neural networks. You will need to study the paper "Backpropagation through time: What it does and how to do it", Paul J. Werbos, Proceedings of the IEEE, volume 78, number 10, 1990. This paper can be accessed via Infolinx and then going to the IEEE Digital Library. You need to implement in Matlab the pseudocode given for backpropagation and backpropagation through time in pages 1552-1554 and 1555-1556. You should apply your implementations in the IRIS data set classification problem which can be found at: http://archive.ics.uci.edu/ml/datasets/Iris.

You should try to finish as much of the material as possible although you will not be asked about any of this material in the examination.

Although most of the time you will be working by yourself, you will have the opportunity to get feedback of how well you are doing and ask questions regarding the elements you study, by seeking the advise and help of your lecturer.


Last modified: Thu Mar 26 17:11:24 GMT 2009